Você está na página 1de 129

SECTION 3

METHANOL PRODUCTION
THE DEVELOPMENTOF LIQUID PHASE METHANOL PROCESS:
AN UPDATE

T. R. Tsao and P. Rao

Air Products and Chemicals, Inc.


ABSTRACT

Air Products and Chemicals, Inc., with the support of the U.S. Department of
Energy (DOE), began a research and development project in September, 1981 as
part of DOE's indirect liquefaction program to further develop the Liquid Phase
Methanol (LPMEOH*) Process at a Process Development Unit (PDU) scale. Chem
Systems Inc., the inventor of the process, is the key subcontractor in the
program. Industrial cost-sharing participants have been Air Products, the
Electric Power Research Institute, and Fluor Engineers, Inc.

During the past wear, a 40-day continuous operation with CO-rich gas
(H2/C0~0,69) was accomplished in the LaPorte PDU with a 25 wt% slurry. The
operating conditions of this run were similar to the one reported last year at
this conference. In this run, catalyst a c t i v i t y and a c t i v i t y maintenance were
excellent, comparable to performance established in bench-scale reactors.
Approximately IB6 metric tons of methanol were produced with a methanol purity
of 96 percent. The PDU on-stream factor was 97 percent. The success of this
run was a major milestone in the development of the LPMEOHtechnology.

A second PDU run with a more concentrated catalyst slurry was also performed.
lhe catalyst was successfully activated at the high slurry concentration. High
methanol production, 7 TPD, was achieved with the CO-rich feed, although the
methanol productivity of the catalyst was lower than expected. The run was i

accomplished with a 100% on-stream factor. There were no operational problems


and catalyst entrainment was modest.
Laboratory programs contributed to the development of i n - s i t u catalyst i
reduction techniques that were successfully used at LaPorte. Based on
autoclave studies and PDU performance, target KLa values were developed for
consideration of future reactor modifications. In addition, studies were
conducted on the effect of nickel and iron carbonyl on methanol catalyst
activity, and the desired levels of C02 in CO-rich gas were identified.

LPMEOH technology is reaching development milestones. Additional PDU


operations are planned, and research programs to meet key technical challenges
are in place. A program has been proposed for a semi-works demonstration
plant at lVA's gasifier site in Muscle Shoals, Alabama.

*A trademark of Chem Systems, Inc.

3-3
I. INTRODUCTION

The reactions of hydrogen and carbon oxides to methanol are very exothermic
(Figure I). High pressure and low temperature favor the reaction equilibrium
in the direction of methanol formation. Early methanol synthesis processes
generally opeFated at pressures of 270-370 atmospheres (4000-5500 psi) and
temperatures oi 340-400°C (650-750°F) with a zinc-chromium catalyst. With the
development of copper-based methanol synthesis catalysts, the operating
conditions were moderated considerably to pressures of approximately 50-I00 atm
(750-I000 psi) and temperatures of 220-270°C (430-520°F).

The most d i l f i c u l t design problem of the methanol synthesis process has always
been removing the heat of reaction while maintaining close temperature control
to achieve optimum selectivity and reaction rate. Catalyst l i f e is seriously
reduced by higher temperatures. In conventional gas-phase reactors, injection
of cool unreacted gas at stages in the catalyst bed or internal cooling
surfaces are employed to provide temperature control. However, these schemes
have been developed for diluted syngas which yields low conversion per pass.

lhe Liquid Phase Methanol (LPMEOH*) process invented by Chem Systems Inc.
dilfers significantly from conventional gas-phase processes in the method
of removing [he heat of reaction. T h i s process utilizes a heterogenous
catalyst lluidized or entrained by a circulating inert hydrocarbon liquid,
usually a mineral o i l . The presence of this liquid serves to control the
reaction temperature much better than in gas-phase processes, allowing a higher
conversion per pass while permitting recovery of the heat of reaction. In
addition, laboratory and Process Development Unit (PDU) tests to date show
LPMEOH technology particularly suited to coal-derived synthesis gas rich in
caFbon monoxide, l hese capabilities make the LPMEOHprocess a potentially
lower-cost conversion route to methanol, especially when methanol coproduction
is added to a coal-based, integrated gasification combined cycle (IGCC) power
plant. For a modest increase in the capital cost and complexity of an IGCC
plant, the methanol coproduction scheme produces a storable liquid fuel in
parallel with electric power production, providing a significant turndown and
peak-load capability for the IGCC plant.

Chem Systems conceived the concept of liquid-phase methanol synthesis in the


mid-1970's. Early research was done on the ebullated-bed reactor, using
relatively large (3-6 mm) catalyst particles fluidized by gas and liquid flow.
lhe development of the liquid phase slurry reactor began at Chem Systems in
1979. lhe i n i t i a l bench-scale work was done in stirred autoclave reactors. At
that scale, the research focused largely on i n t r i n s i c catalyst performance:
catalyst screening, activation, and l i f e tests.

In September 198~, the United States Department of Energy (DOE) awarded a


contract entitled "Liquid Phase Methanol Process Development Unit:

*A trademark of Chem Systems Inc.

3-5
lll I

I.~(.11 I.~( I,,,1, l l p e l a l l , ~ n and Kllpp,~rf R t I l d i e ~ " l~hich ~'a~ the f i r s t phase of a
tJl~hli,Z.J ,tlz~l,',l ,tt luxlhz'v d~,velL~plng tlze LI'MLOtt p l ' o c e ~ Ill a representative
engineering-scale PDU. A second contract began in July 1985. Air Products and
Chemicals, Inc. is the prime contractor providing overall program management

and has been responsible for engineering design, procurement, construction, and
operation of the PDU. ChemSystems is performing as the key subcontractor in
the program. Cost-sharing has been provided by Air Products, the Electric
Power Research Institute (EPRI), and Fluor Engineers, Inc.
I

In this program, a DOE-owned skid-mounted p i l o t plant was disassembled and


equipment components renovated. The unit was transferred from Chicago,
I l l i n o i s to Air Products' LaPorte, Texas f a c i l i t y , refurbished, and rebuilt for
I service as the LPMEOH PDU. Synthesis feed gas from the f a c i l i t y is used to
I test the unit. The LaPorte LPMEOH PDU design provides for a liquid-fluidized
(ebullated-bed) mode of reactor operation and a liquid-entrained (slurry) mode
of reactor operation.

A total of five major runs have been conducted at LaPorte since its
commissioning in March 1984. The results of the f i r s t three runs made in 1984
were reported in the 1984 and 1985 EPRI Contractors' Conferences. The
operation and results of the latest LaPorte PDU runs (E-3 and E-4) completed in
1985 are discussed in this paper.

The development of the LPMEOHprocess is supported by extensive laboratory


programs funded by both DOE and EPRI, which include catalyst screening,
bench-scale tests, fundamental modeling, poisons studies, C02 effect on
methanol productivity, alternate liquid screening, slurry c r i t e r i a study,
and the effect of in-situ reduction conditions on catalyst a c t i v i t y . Recent
results of the research and development programs are presented in this paper.

3-6
2. LAPORTE PDU OPERATION

LaPorte PDU Description

The primary Function of the LaPorte PDU is to acquire data at a small,


representative engineering scale for testing the f e a s i b i l i t y of the LPMEOH
process. Thus, the PDU was designed with the capability of generating and
collecting plant data over a wide range of operating conditions. The range of
operating variables chosen for design is shown in Table I . As w i l l be apparent
later, some of the design ranges were exceeded in actual operation.

lhe principal reactor feed gas compositions considered during design were:

Balanced Type (Table 2), in which the hydrogen and carbon oxide
concentrations are approximately stoichiometrically balanced in
order to achieve an "all4methanol '' product.

CO-rich Type (Table 3), in which the hydrogen and carbon oxide
concentrations are not stoichiometrically balanced, but are
representative of synthesis gases from modern coal gasifiers.
These gases are suitable for once-through methanol synthesis
in an IGCC flowsheet configured to make electric power and
coproduct methanol.

lhe LaPorte PDU was designed to test both the ebullated-bed mode and the
slurry mode of operation. A unified design concept was used so that a common
reactor and PDU system could accommodate both operating modes. Equipment,
instrumentation, and valving specifications included consi'deration of both
modes of operation from the start of the design effort. As a result, the
LaPorte PDU can be switched from ebullated-bed to slurry operation without
equipment or piping alterations.

lhe d i f f e r e n t reactor feed gas compositions are blended from H2, CO, N2, and
CH4 supplied by the adjacent syngas f a c i l i t y . Carbon dioxide is trucked into
the plant as a l i q u i d and stored o n - s i t e . Since only a portion of the reactor
feed is converted per pass, the unconverted synthesis gas is recycled and mixed
with fresh makeup gas. The makeup gas is blended so t h a t the reactor feed
(makeup plus recycle) simulates e i t h e r the balanced or CO-rich gas type.
Recycling the unconverted synthesis gas reduces gas consumption by 70 percent.

A simplified process flowsheet for the LaPorte PDU is shown in Figure 2. The
makeup synthesis gas is compressed to the reactor pressure (3,500-6,300 kPa,
500-900 psig) by the feed compressor. The compressed makeup and recycle gases
are mixed and preheated in the feed/product exchanger before being fed into
the methanol reactor, the inert hydrocarbon liquid or slurry that circulates
through the reactor is separated from the unconverted synthesis gas and
methanol product vapor in the primary V/L separator, and recirculated to the
reactor through the slurry heat exchanger. The circulating liquid or slurry
can be heated or cooled in the slurry exchanger to maintain a constant reactor
temperature, depending upon the level of conversion, system heat losses, and

3-7
the rate of cold seal flush required by the slurry pump. A u t i l i t y oil system
provides the heating or cooling duty to the slurry exchanger.

lhe unconverted synthesis gas/product methanol stream leaving the primary V/L
separator is cooled against incoming feed gas and condensed oil is separated in
the secondary V/L separator. The uncondensed vapor is further cooled in the
product coo)er. Condensed methanol is then separated from the synthesis gas
and additional condensed oil before routing to product storage. A small purge
stream is sent to flare. The bulk of the unconverted synthesis gas is
compressed and returned to the front end of the PDU. Additional systems are
present to activate the catalyst, provide seal flush to the slurry pump, and
mix catalyst slurry for the liquid-entrained mode of operation.

l aPorte PDUOperating Results

A total of five major runs have been conducted at the LaPorte PDU since
commissioning in March 1984. A summary of these campaigns is presented
in fable 4. lhe results of Runs F-l, E-l and E-2, including two phase gas
holdup studies, were reported in the 1984 and 1985 EPRI Contractors'
Conferences. (1,2)

The f i r s t PDU run (F-l) was a lO-day shakedown r u n . Operation of the PDU was
smooth, and the mechanical i n t e g r i t y and process f l e x i b i l i t y of the unit were
demonstrated. Up to 8 IPD of methanol was produced. The second PDU run (E-l)
was a 40day continuous run on CO-rich synthesis gas (H2/C0=0.7). Stable
operation was achieved but a slow, continuous decline in a c t i v i t y was observed,
in excess of that anticipated from isothermal laboratory autoclaves. The
accumulation of trace poisons on the catalyst was the major cause of this loss
of a c t i v i t y (I.1% per day). A third PDU run (E-2) was conducted using a
commercially available catalyst powder at high slurry concentration. In-situ
reduction was performed. The plant operated well mechanically, providing
valuable experience for the operations and engineering staff on handling
hiqh viscosity catalyst slurries. Methanol productivity was below the values
predicted from previous laboratory results.

Out of a supporting laboratory program funded by EPRI, a series of tests were


conducted and i t was found that inadequate catalyst activation at LaPorte was a
contributor to the off-performance at the high solids loading. (3) Changes in
the reduction procedure were identified to remedy this problem. Mass transfer
resistance may also have contributed to the reduced catalyst performance during
Run E-2, but i t s existence was masked by the inadequate catalyst activation.
Another hiqh slurry concentration PDU run with properly activated catalyst was
deemed necessary to determine the reactor productivity at high slurry
concentrations.

Analysis of the results of the 1984 operating program indicated that selective
upqrading of materials of construction of the LaPorte PDU would lead to
lower levels of trace contaminants. Process improvements which would increase
the data gathering capability were also specified. As a result, modifications
were made to the [aPorte PDU during early 1985. New equipment was installed to
improve the measurement of slurry concentration and methanol product flow.
Also, selected vessels and piping were replaced or modified in order to reduce
the levels of trace catalyst poisons, primarily iron and nickel carbonyls. A
chemical cleaning program was also undertaken to remove residual contaminants.

3-8
Upon completion of these a c t i v i t i e s , a second 40-day a c t i v i t y maintenance test
(Run E-3) using CO-rich gas and a 25 wt% catalyst slurry was conducted in
May-June 1985. This was followed by a more concentrated slurry test
(Run E-4), which was performed under the new contract with the DOE for a
second phase of LPMEOHdevelopment. The results of these latest LaPorte PDU
runs made in 1985 are presented in t h i s paper.

LaPorte PDU Run E-3

The fourth LaPorte PDU Run E-3 took place in May-June 1985. The primary
objective of this 40-day operation was to demonstrate improved a c t i v i t y
maintenance of the LPMEOHprocess with CO-rich gas, with trace contaminants
eliminated and using catalyst powder which had been reduced i n - s i t u (Table 5).
A fresh batch of the same catalyst powder used in Run E-2 was slurried in
oil and transferred to the reactor system. I n - s i t u reduction of the 25 wt%
slurry was then performed. Hydrogen consumption during reduction is the
prime indicator of the progress of catalyst reduction. The hydrogen uptake
matched satisfactory autoclave reduction runs (Figure 3). This indicated that
a successful i n - s i t u reduction had been accomplished in the PDU. CO-rich
synthesis gas was then brought into the PDU and the reactor conditions were
adjusted to the f i r s t condition listed in Table 6 (E-3A). Two operating points
were tested over the 40 days of operation. Case E-3A, which was a duplication
of the a c t i v i t y maintenance condition of Run E-l, was held for the i n i t i a l
94 hours to establish a baseline catalyst a c t i v i t y . The second case (E-3B) was
a brief test at a lower reactor temperature of 225°C (437°F). Reactor
conditions were then returned to 250°C (482°F) for the remainder of the run
(E-3C) to determine the a c t i v i t y maintenance characteristics of the catalyst.

Highlights of the LaPorte PDU operation during Run E-3 are presented in
fable 7. Overall, the PDU performed well, achieving a 97% on-stream factor and
producing over 186 metric tons of crude methanol. The major fraction of the
downLime (34 hours) was due to an electrical f a u l t in the motor for the feed
compressor. A replacement motor was located and installed, and synthesis gas
was brought back into the PDU. The outage, though unplanned, demonstrated
the a b i l i t y to maintain catalyst a c t i v i t y through a temporary plant shutdown.
The run ended on 13 June after the planned 40 days of operation.

Figure 4 shows the CO conversion and methanol productivity as a function of


time on synthesis gas for Run E - 3 . The autoclave prediction is also presented
for comparison. The LaPorte PDU data have been normalized to a space velocity
of lO,O00 I/hr-kg to provide a common basis of comparison between the PDU data
and the laboratory results. I t is seen that the PDU performance is comparable
to the laboratory predictions throughout the duration of the r u n .

lhe LaPorte PDU data for the f i r s t several days e x h i b i t a high a c t i v i t y that
does not f i t the linear decline in a c t i v i t y observed for the remainder of the
FUn. When these i n i t i a l hyperactivity points are excluded, a 0.28% per day
decline in methanol productivity is seen over the operation of Run E-3. The
significant improvement over the l.l% per day decline observed in Run E-l is
believed to be a direct result of removing catalyst poisons by chemical
cleaning and the metallurgical upgrade performed before the run.

In ~iguFe 5, the a c t i v i t y maintenance history based on cumulative methanol


pFoducLion is depicted for Run E-3 and for two earlier runs - LaPorte PDU
Run E I and a 2,267-hour laboratory autoclave run completed in October 1983.

3-9
The laboratory result represented the previous best performance with
CO rich gas. Figure 5 illustrates that the deactivation rate for Run E-l is
approximately a factor of 4 greater than that for the autoclave test. However,
after the completion of the metallurgical upgrade and chemical cleaning of
the PDU, Run E-3 yielded a deactivation rate approaching that in the
autoclave.

fable 8 compares the results of catalyst analyses from Runs E-1 and E-3. I t is
evident that there was essentially no increase in the levels of trace catalyst
poisons in Run E-3. The significant improvement over the previous run data
(Run E-l) verified the effectiveness of the metallurgical upgrade and chemical
cleaning, lhe achievements of this run are summarized in Table 9.

LaPorte PDU Run E-4

lhe f i f t h LaPorte PDU Run E-4 was a lO-day run conducted during the summer of
1985 to demonstrate in-situ reduction of a high slurry concentration and to
obtain performance data with high solids loadings. T h i s run was a repeat of
Run E-2 which had less than expected performance due to unsatisfactory i n - s i t u
reduction, Catalyst powder was slurried to a concentration of 43 wt% in the
slurry prep tank and transferred to the reactor system. Improved catalyst
reduction techniques resulting from laboratory programs were followed. Total
hydrogen consumption agreed well with autoclave results, indicating a
successful in-situ reduction.

After reduction, the slurry was concentrated to 47 wt% (as oxide) during the
f i r s t few hours under CO-rich gas. The reactor was maintained at 5,300 kPa
(750 psig), 250°C (482°F), and with a gas superficial velocity of 15 cm/sec
(0.5 ft/sec). PDU performance started well but methanol productivity degraded
rapidly and a solids concentration gradient appeared in the reactor. Both
liquid and gas flow rates were increased with no apparent effect on methanol
productivity and the solids concentration gradient. The slurry was
subsequent]y diluted to 40 wt% and later to 34 wt%. With each d i l u t i o n , the
reactor performance improved, approaching that of the autoclave. A uniform
solids concentration was restored at the 40 wt% slurry loading. At 34 wt%
slurry loading, the methanol productivity improved to a level equivalent to 85%
of the autoclave performance, producing 6.9 TPD methanol with CO-rich gas.
Stable operation was maintained at this condition for four days with no
apparent a c t i v i t y decline. The mechanical performance of the LaPorte PDU was
excellent during this run, achieving a I00% on-stream factor. There were no
problems with slurry pumping or plugging, and catalyst entrainment was modest.
Approximately 68 tons of crude methanol with a 95% purity were produced. The
highlights of Run E-4 are summarized in lable lO.

Subsequent tests in well-mixed autoclaves on slurry samples taken directly from


the LaPorte run verified that the i n t r i n s i c a c t i v i t y of the catalyst was
normal, lherefore, i t is believed that the performance of the PDU reactor was
probably hindered by either mass transfer limitations or inadequate solids/gas
mixing at the higher slurry concentrations. To f u l l y exploit the potential of
high (>40%) slurry operation, engineering studies on alternate reactor systems
as well as research work on alternate liquid media and a better understanding
of good slurry behavior, are being carried out in Phase I I of the DOE-sponsored
Liquid Phase Methanol program.

3-10
r
3. RESEARCH RESULTS

The on-going laboratory e f f o r t to support the development of the LPMEOHprocess


has the follew~-: e , : ~ r ~ c~:%~,e~:

(a) F i r s t , i t is desired to further the fundamental understanding of the process


and catalyst. Examples of past work include studies of i n - s i t u reduction
and defining surface properties of properly activated catalysts (3).

(b) A second objective is to conduct systematic research towards further


improvements in performance of the LPMEOH reactor.

(c) A third objective is to provide technical support during start-ups and


I operation of the LaPorte PDU. Examples include poisons monitoring of
species such as carbonyls, chlorides and hydrogen sulfide, catalyst
q u a l i f i c a t i o n prior to start-up, and catalyst characterization during
operation. When required, short-term laboratory programs are i n s t i t u t e d to
troubleshoot a specific problem.

lhe results and conclusions in three tasks are presented here. These are:
( i ) lhe effect of carbonyls (Ni and Fe) on catalyst a c t i v i t y and properties;
( i f ) lhe effect of gas composition (primarily C02) on liquid phase operation; and
( i i i ) Studies on alternate liquids and mass transfer limitations.

DESCRIPIION OF EQUIPMENT

Liquid phase operations in the laboratory are conducted in stirred autoclaves. A


simplified diagram of a system using a l - l i t e r autoclave is shown in Figure 6.
lhe system is capable of operation at high pressure and temperature with a
variety oi prebiended gases. Gases are stored in large cylinders on t r a i l e r s and
are f i r s t passed through adsorbent guard beds to remove poisons prior to delivery
to the autoclave. The system is housed in a walk-in hood with CO alarms and is
completely automated for safe, attended and unattended operation. There are two
additional autoclave systems, each of 300 cc volume that are similar to the l
l i t e r autoclave system. Slurry samples can be withdrawn during methanol
synthesis. A dedicated GC provides the necessary analytical capability for the
calculation of mass balances and the reporting of results. In addition, there
are various other gas phase test units in support of the autoclaves. Analytical
resources provide necessary analyses such as ESCA/AUGER, X-Ray Diffraction,
Atomic Absorption Spectroscopy, BEI surface area measurements, and other
state-of-the-art measurements.

EFFECT OF NICKEL AND IRON CARBONYLS

lhe effect of nickel and iron carbonyl was studied in two separate autoclave
runs. Methano] catalyst was slurried with hydrocarbon liquid Freezene-lO0 oil
and loaded into a 300 cc autoclave. Standard i n - s i t u activation procedures were
used and the autoc]ave was run on a poison-free, CO-rich synthesis gas for 90
hours. Autoclave conditions were 250°C (482°F), 5,300 kPa (750 psig), at a

t
3-11
nominal space velocity of 5000 SL/hr-kg. Stable and satisfactory operation was
confirmed and injection was begun of a gas stream containing nickel carbonyl.
lhe combined feed to the autoclave contained between 0.5 and l ppmv of Ni(CO)4.

lhe plot of methanol productivity with time is shown in Figure 7. For the f i r s t
90 hours when no nickel carbonyl was being injected, the performance was
stable and for the conditions tested, agreed well with the expected performance.
Upon the injection of nickel carbonyl, the methanol productivity began to
decline. Operation was terminated after about 80 hours of operation under
poisoning conditions. Catalyst samples were withdrawn at various times during
the run and analyzed. These results w i l l be discussed later in t h i s paper.

A second autoclave run was conducted to study the effect of iron carbonyl on
catalyst performance. Once again, the catalyst was i n - s i t u reduced, and operated
with poison-free, CO-rich gas for a s u f f i c i e n t run-in period, in t h i s case 120
hours, lhe catalyst performance was stable and agreed well with the
expectations. Upon injection of iron carbonyl, the a c t i v i t y began to decline as
shown in Figure 8. The run was terminated after about 120 hours on
poison-containing gas. As in the nickel carbonyl run, catalyst samples were
taken at various times.

The analyses of the catalyst samples are summarized in Tables l ] and ]2 f o r


the nickel and iron carbonyl runs, respectively. These results confirm that the
catalysts were absorbing nickel or iron during the poisoning experiments.
However, within experimental error, no effect could be discerned either
in crystal size measurements of Cu and ZnO or BET surface area. The ESCA/AUGER
analyses proved to be inconclusive and no Ni or Fe was detected on the surface,
though t h e i r presence in the bulk solid was confirmed by AAS. I t is possible
that the washing operation used in removing the catalyst from the slurrying o i l
removes the Ni and Fe from the surface.

Relative a c t i v i t y decline as a function of the nickel and iron content is shown


in Figure 9. I t is interesting to note that for all practical purposes, nickel
and iron appear to be equivalent in t h e i r a b i l i t y to destroy catalyst a c t i v i t y .
I t is also interesting to note that these data indicate a l e v e l l i n g - o f f effect at
about 500 ppm of nickel or iron. Longer runs would have to be conducted to
confirm this with a degree of certainty. Also shown in Figure 9 is the relative
a c t i v i t y decline as a function of Fe and Ni in the catalyst from the LaPorte PDU
run E-I. This run used a d i f f e r e n t catalyst with a d i f f e r e n t composition. The
LaPorte data also cover a longer period of time than the autoclave studies. The
data in Figure 9 may indicate that d i f f e r e n t catalysts d i f f e r in t h e i r a b i l i t y to
withstand carbonyl poisoning.

ALIERNAIE LIQUIDS FOR THE LPMEOH PROCESS

The study on alternate liquid candidates was conducted to find a satisfactory


substitute For Witco Freezene-lO0 o i l . This included developing c r i t e r i a for a
satisfactory l i q u i d , surveying commercially available liquids and selecting a few
for analyses and autoclave tests.

A satisfactory liquid for the LPMEOHprocess must meet certain requirements. I t


must be inert and not react with the feed gases or catalyst. I t must be within
the suitable range of properties such as v i s c o s i t y , density, gas s o l u b i l i t y , and
surface tension in order to permit satisfactory catalyst suspension and
gas bubble formation. The liquid must not have components that can poison the
catalyst such as trace metals, halogens, sulfur compounds and unsaturates or

3-12
unstable compounds. The liquid should permit in-situ reduction and i t s boiling
point must be high enough so as not to have excessive vapor pressure at operating
temperature. Finally, i t should be commercially available at a reasonable cost.
Many liquids were considered, and, based on physical and chemical properties,
candidates were selected for further testing. The candidates included Exxon
200 which is a silicone-based o i l , Witco-70, Witco LP-150, and
HI-43, Dow Corning test program included catalyst activation, a short-term
Amoco l8 USP. The owed by a s t i r r e r RPM study to distinguish liquids of superior
a c t i v i t y check l o l l
mass t r a n s f e r c a p a b i l i t i e s . The a c t i v i t y r e s u l t s are p l o t t e d in Figure 10.
Satisfactory performance is indicated by data f a l l i n g on the standard performance
curve. Clearly, successful candidates were Witco-70 and LP-150. The cause of
the f a i l u r e of Exxon H~-43 is a t t r i b u t e d to a 2 to 3% level of aromatics which
may have poisoned the c a t a l y s t . The c a t a l y s t from the Dow Corning o i l run showed
high levels of Si contamination which may have been responsible f o r i t s poor
performance, lhe Amoco-18 USP o i l did not perform s a t i s f a c t o r i l y .

Results from the RPM study with successful candidates are shown in Figure 11.
Both Witco-lO and LP-150 are equivalent to Freezene-lO0. The Freezene-lO0 curve
was developed at a s l i g h t l y higher space v e l o c i t y , which is the reason f o r
the methanol p r o d u c t i v i t y l e v e l l i n g off higher at the higher RPMs. The regime of
mass t r a n s f e r control in the autoclave at the conditions used, is i d e n t i f i e d to
beg~n below abouL 700 RPM.
lhe data from Figure l l were used in conjunction with a kinetics/mass t r a n s f e r
model to calculaLe KLa requirements. The Yagi and Yoshida (4) c o r r e l a t i o n f o r

autoclaves was u~ed with s o l u b i l i t y data from Matsumato and Satterfi eld.(5) With
the data base from laboratory and LaPorte PDU operations, i t is now possible to
estimate the desirable KLa values for future improvements in reactor designs both
at LaPorLe and in larger demonstration units.

EFFECT OF GAS COMPOSITION (CO~


l h i s task was undertaken to determine the e f f e c t of CO: in CO-rich gas on
methanol c a t a l y s t a c t i v i t y and p r o p e r t i e s . The gas compositions used in t h i s
study ace l i s t e d in lable 13. lhe reference was CO-rich syngas composition
containing 13% CO:. Gases B, with 8% C02, and C, with 4% C02, were picked to
study the impact of C02 at constant p a r t i a l pressures of CO and H2. Gas D was
used to see how much methanol could be produced over a stoichiometric CO:/H2
gas. Reaction conditions were fixed at 250°C (4 B2°F) and 5,300 kPa (750 psig).
lhe catalyst was slurried with oil and reduced in-situ. All synthesis results
are based on a week of stable performance, and are summarized in Table 14 and
shown in Figure 12. Based on these data, i t is concluded that for a CO-rich
feed, optimum performance is achieved with a C02 content somewhat higher than 13%
in the inlet gas, although there is not much change in performance once the C02
content exceeds 7-8%. A more rigorous analyses of the data is the subject of
ongoing work. Ibis effect w i l l be further studied in the laboratory as part of a
future program. Surface analysis data on the catalysts have shown no obvious
eft errs of the variable C02 levels.

3-13
4. SUMMARY

LaPorte PDU operations over the past two years have contributed significantly
toward demonstrating LPMEOHtechnology at a representative engineering scale. The
PDU has accumulated over 2,500 hours of synthesis gas operation with an on-stream
factor of 96-I00 percent. The f e a s i b i l i t y of operating the liquid-entrained
system with a 25 wt% catalyst slurry for an extended period of time and converting
a portion of a CO-rich synthesis gas to methanol with low catalyst deactivation is
a notable achievement. The a b i l i t y to activate methanol synthesis catalyst
powders in an inert liquid at high concentrations is also noteworthy. Methanol
production levels as high as 8 TPD for balanced gas feed and 7 TPD for CO-rich gas
feed were achieved; the purity of the methanol product from CO-rich gas is
consistently higher than 96 wt%, a good fuel grade quality.

The extensive supporting research programs have furthered the understanding of the
LPMEOH process and catalyst performance and provided technical support during
LaPorte PDU operations. The research work has solved key technical problems
identified during the PDU operation. A modified in-situ activation procedure for
a concentrated slurry was developed in the laboratory and successfully practiced
at LaPorte. The poisoning impact of iron and nickel carbonyl was quantified. The
data on the desirable C02 content in a CO-rich feed was determined for future
commercial operation.

3-15
5. FUTUREWORK

Work is in progress to evaluate modified reactor designs to improve reactor


productivity. Improved methods for poisons detection and control must be
developed for gases from coal gasifiers, and are being studied. Activity
maintenance through temperature programming w i l l be practiced in the autoclave.
A systematic study is continuing on slurry properties and behavior in relation
to catalyst activation and a c t i v i t y . Further work on the effect of C02 is
planned. Improvements in the kinetic model and the evaluation of other catalysts
will be conducted. Additional LaPorte PDU runs are planned for the demonstration
of activity maintenance with catalyst addition and withdrawal, as well as
improved reactor design. Longer term l i f e runs at the PDU level are
contemplated.

From LaPorte, i t is anticipated that the LPMEOHtechnology w i l l advance to a


semi-works development/demonstration scale. A proposal has been submitted to DOE
under the Clean Coal Technology Program for a 35 TPD demonstration unit with the
host site being TVA'S Muscle Shoals, Alabama f a c i l i t y . Clean CO-rich synthesis
gas from the lexaco coal gasifier w i l l be available as once-through feed gas to
the LPMEOHreactor.

In sun~ary, the LPMEOHprocess is reaching development milestones. The results


to date are encouraging, although some technical challenges remain.
The technology is positioned for advancement to a demonstration f a c i l i t y in the
near future.

3-17
6. REFERENCES

l. J. KIosek and R. L. Mednick, "Progress in Liquid Phase Methanol


Development", presented at the 9th Annual EPRI Contractors' Conference on
Clean Liquid and Solid Fuels, 8-I0 May 1984.

. J. Klosek, D. M. Brown and R. L. Mednick, "Status of the LaPorte Methanol


PDU", presented at the lOth Annual EPRI Contractors' Conference on Clean
Liquid and Solid Fuels, 23-25 April 1985.

3. D. M. Brown, T. H. Hsiung, P. Rao and M. I . Greene, "Catalyst A c t i v i t y and


Life in Liquid Phase Methanol", presented at the lOth Annual EPRI
Contractors' Conference on Clean Liquid and Solid Fuels, 23-25 April 1985.

4. H. Yagi and F. Yoshida, "Gas Absorption by Newtonian and Non-Newtonian


Fluids in Sparged Agitated Vessels," I&EC Proc. Des. Dev. 14 (4), 1975.

. D. K. Matsumato and C. N. S a t t e r f i e l d , " S o l u b i l i t y of Hydrogen and Carbon


Monoxide in Selected Nonaqueous Liquids," I&EC Proc. Des. Dev. 24 (4),
1985.

3-19
RANGE OF OPERATING VARIABLES FOR
LAPORTE PDU
MINIMUM "NORMAL" MAXIMUM

REACTOR PRESSURE, KPA 3,500 5,300 6,300


REACTOR TEMPERATURE, °C 220 250 270
LIQUID-FLUIDIZED SPACE
!
PO
O
VELOCITY, LITER/HR-KG CAT. 1,000 2,500 4,000
LIQUID-ENTRAINED SPACE
VELOCITY, LITER/HR-KG CAT. 2,000 6,000 10,000
LIQUID-FLUIDIZED CATALYST
LOADING, SETTLED BED HEIGHT, FT 5 7 7
LIQUID-ENTRAINED CATALYST
LOADING, WT. % 10 20 33

NOTE: SPACE VELOCITY BASED ON STANDARD LITERS (0°C, 14.7 PSIA), KG OF OXIDE CATALYST, AND
ZERO GAS HOLDUP IN REACTOR,

Tabl e 1
LAPORTE PDU PRINCIPAL
FEED GAS COMPOSITION
ALL-METHANOL PRODUCT

BALANCED TYPE
REACTOR FEED

H2 54.9 MOLE %
CO 18.8
002 4.9
OH4, 02H6 2.1
N2, Ar, INERTS 19.3
C.J
!
TOTAL 100.0
PO
H2/CO 2.92
H2 2.10
(CO + 1.5 002)
(HI2 -- 0 0 2 ) 2.11
(CO + 002)

RECYCLE CONVENTIONAL
MEOH SYNTHESIS

,~ PURGE • FRESH FEED SHIFTED


• CO2 REMOVED
H_~2 ~--, 2 ~
CO MEOH
Table 2
LAPORTE PDU PRINCIPAL
FEED GAS C O M P O S I T I O N
C O P R O D U C T M E T H A N O L + ELECTRIC POWER

CO.RICH TYPE
REACTOR FEED

H2 34.8 MOLE %
CO 51.2
002 13.1
CH4, 02H6 0.1
C~
!
N2, Ar, INERTS 0.8
r~
r~ TOTAL 100.0
H2/CO 0.68
H2 0.49
(CO + 1.5 CO2)
(H..2 - CO2) 0.34
(CO + CO2)

FUEL ONCE-THROUGH MEOH


> TO GAS SYNTHESIS, IGCC FLOWSHEET
TURBINE
H2 (ELECTRIC
c--b-o.7 /
POWER) • NO SHIFT
MEOH • NO CO2 REMOVAL
Table 3
LAPORTE PDU OPERATION SUMMARY
RUN OPERATION CATALYST H O U R SON

F-1 MAR 1984 SHAKE- EBULLATED --.--~ EXTRUDATES 248


DOWN HYBRID'-'~
ENTRAINED
964
EBULLATED "-'~ EXTRUDATES
E-1 APR/MAY 1984 ACTIVITY HYBRID----~
MAINT.
C,J
ENTRAINED
I
145
PO
CO
ENTRAINED POWDER
E-2 JUN 1984 HIGH SLURRY
CONC., HIGH
THROUGHPUT
948
ENTRAINED POWDER
E-3 MAY/JUN 1985 ACTIVITY
MAINT.
231
ENTRAINED POWDER
E-4 JUL 1985 HIGH SLURRY
CONC., HIGH
THROUGHPUT 2536

Table 4
OBJECTIVES OF LAPORTE PDU RUN E I 3

1. D E M O N S T R A T E IN I S I T U REDUCTION AT
CO
!
LAPORTE PDU S C A L E
P,O

2. D E M O N S T R A T E IMPROVED A C T I V I T Y
M A I N T E N A N C E A F T E R METALLURGICAL
CHANGES AND C H E M I C A L CLEANING

Table 5

IIII II II IIII I II1~ ~ , ~ I _ J, I I I I I I . . . . . . . . . . . . . . . . . . . . . . . . .


LAPORTE PDU RUN E - 3 O P E R A T I N G C O N D I T I O N S
(3 M A Y - 13 JUNE 1 9 8 5 )

CATALYST: POWDER
GAS TYPE: CO-RICH
REACTOR PRESSURE: 5,300 KPA
SUPERFICIAL LIQUID VELOCITY: 4.9 CM/S
~O
!
U'I
SUPERFICIAL
GAS VEL. SPACE VEL. SLURRY CONC. HRS. AT
CASE T (°C) (CM/S) (L/HR-KG) (WT% OXIDE) COND.
E-3A 250 9.5 10,000 28 94
E-3B 225 8.8 11,300 25 23
E-3C 250 9.5 10,000 28 831
948

Table 6
LAPORTE PDU HIGHLIGHTS - RUN E-3

• SMOOTH CATALYST LOADING, MIXING, AND SLURRY


TRANSFER

• CONTINUOUS SMOOTH OPERATION OF SLURRY


CIRCULATION PUMP - ALMOST 100% ON-STREAM TIME

C~
!
• ACHIEVED 97% OVERALL ON-STREAM TIME
C~

• 34 HOUR OUTAGE DUE TO COMPRESSOR MOTOR PROBLEM

• DEMONSTRATED ABILITY TO MAINTAIN CATALYST


ACTIVITY AFTER EXTENDED PDU SHUTDOWN

• LOW CATALYST CARRYOVER

e PRODUCED 186 METRIC TONS METHANOL WITH 96% PURITY

Table 7
LAPORTE PDU:
CATALYST ANALYSES FOR RUNS E-1 AND E - 3

RUN E- 1 RUN E- 3
I I IIII

L,o

PO
! HRS. ON Fe Ni HRS. ON Fe Ni
..,.,.j
SYNGAS (PPMW) (PPMW) SYNGAS (PPMW) (PPMW)

0 165 42 0 68 37

964 394 137 942 67 26

Table 8
LAPORTE PDU: ACHIEVEMENTS OF RUN E - 3

• SUCCESSFUL IN-SITU REDUCTION OF A 25 WT%


CATALYST SLURRY AT LAPORTE PDU SCALE
I

CO • ELIMINATION OF C A T A L Y S T POISON ACCUMULATION

• OPERATION WITH CO-RICH GAS WITH LESS THAN


0.3% / DAY CATALYST DEACTIVATION

• 97% ON-STREAM FACTOR

Table 9
LAPORTE PDU: S U M M A R Y OF RUN E - 4
(JULY 1985)

• SMOOTH PDU OPERATION AT HIGH SLURRY


CONCENTRATION
C,J
!
P~
• IMPROVED IN-SITU REDUCTION TECHNIQUE
SUCCESSFULLY DEMONSTRATED

• HIGH METHANOL PRODUCTION ACHIEVED WITH


CO- RICH GAS

• 100% ON-STREAM FACTOR

Table 10
EFFECT OF Ni(CO) 4 ON METHANOL CATALYST

XRD (A) BET


HRS. ON Ni Fe
Ni(CO)4 (PPMW CAT.) (PPMW CAT.) Cu ZnO (M2/GM)

0 41.5 40.1 87.5 53.8 106.6


(.,o
I
C,J
O
24 151,0 67,8 92.4 67.0 94.2
48 299.0 51.0 87.7 60.0 105.3
54 416.0 60.3 75.7 67.0 92.4
74.5 542.0 60.6 79.2 69.9 97.1
83 712.0 69.1 75.7 67.0 99.2

Table 11

. . . . . . . . . . . . I III I ..I II li . . ]! . . . . . I JR j III I I I . . . . . . J ..... IU L i ._J-._ ~ _1 1 J] " - /117 ] ] ] r - .... : _!1 :]_~_~,EL~ L_ .... ~ - : ~
EFFECT OF Fe(CO)5 ON METHANOL CATALYST

XRD (A)
HRS. ON Ni Fe iiii i ii BET
Fe(CO)5 (PPMW CAT.) (PPMW CAT.) Cu ZnO (M2/GM)

0 87.9 29.0
CO
I

24 65.2 66.8
CO
p_t
74.0 62.0 98.9
48 70.0 103.0 77.4 71.4 105.6
72 74.0 180.0 72.4 64.4 99.2
96 87.0 297.0 75.7 69.4 97.4
120 67.3 452.0 79.2 80.2 101.1

Table 12
SYNGAS COMPOSITION TESTED

COMPOSITION (MOL %)i

SYNGAS H2 CO CO 2 N2
i i I

CO
!
A 35 51 13 1
CO
P~

B 35 51 8 6

C 35 51 4 10

D 65 0 21 14

Table 13
A U T O C L A V E RESULTS ON CO 2 EFFECT

EXPERIMENTAL CONDITIONS
• REACTOR TEMP: 250°C (482°F)
• REACTOR PRESSURE: 5,300 kPa (750 PSIG)

CO
I
CO
C~
AUTOCLAVE PERFORMANCE
MeOH
PRODUCTIVITY CO CO 2
SYNGAS (G-MOL/HR-KG) CONV. (%) CONV. (%)
A 23.2 13.4 m

B 21.6 12.3
C 18.6 10.5 n u

D 5 iN IN
17

Table 14
METHANOL SYNTHESIS

CO -i- 2H2 ~ CH3OH -!- HEAT (39,500 BTU/MOLE)

f,,,O
!
002 -~- 3H2 ~ CH3OH -!- H20 -I- HEAT (22,000 BTU/MOLE)
,,,p=,

TYPICAL REACTION CONDITIONS:

490°F, 1000 PSIG

Figure 1
SIMPLIFIED PROCESS FLOWSHEET
FOR LAPORTE PDU
MISTER~
RECYCLE
' t~ ~ PURGE
RECYCLE COMPRESSOR

,re, _.__.
SYNTHESIS CW
i
EXCH. L-J
-F
~OD.
COOLER( ~'
FEED GA~S
, ~.oo. s~)
REDUCING
GAS TO -,~
I
FLARE
M.I ASSER
0.) L I
!
C~ [P I

REACTOR

LP L-L SEPR.I

UTILITY UTILITY OIL SYSTEM


OIL IN
EXP. TANK

I
TO MeOH
SLURRY COOLER STORAGE
EXCH. HEATER TANK
UTILITY -- ~ PUMP
- -
)IL RETURN

SLURRY
STEAM ~ CIRCULATION OIL CONDEN- FILTER LP OIL
HEATER " ELECTRIC PUMP SATE PUMP RETURN
HEATER PUMP
Figure 2
HYDROGEN UPTAKE AND WATER FORMATION
DURING REDUCTION
2.0 40
LAPORTE PDU RUN E - 3
..I ,',H 2 = 1 . 5 2 S C F / L B
0 A
-,AUTOCLAVE RUN UJ
AH 2 = 1.5 SCF/L B 30
d 1.5 m

I.U
=E .J
I LU
O't 0r~ ~r
z 1.0 H, 20
O F-
ro
L) X
z uJ
u.I O

0
II:
.5
/ /~2o 10

C~
>..
0
5 10 15 20
TIME INTO REDUCTION, HRS.

Figure 3
LAPORTE PDU 4 0 - D A Y RUN PEFORMANCE
MAY / JUNE 1 9 8 5
35 PRODUCTIVITY, CO-RICH GAS
G MOLE MeOH / HR- KG OXIDE 250°C
7 5 0 PSIG
30 m

8 7 , 0 0 0 SCFH
G MOLE • °• , , • , 10,000 L/HR-KG SV

HR-KG 25
! A U T O C L A V E PREDICTION
,...,.j
20
CO CONVERSION, %

15
% ~
• • OO0
• O 0 O0 O 0 O0 O000 000 O 000

10
A U T O C L A V E PREDICTION
I I I I ! I I I
5
0 5 10 15 20 25 30 35 40
TIME ON S Y N G A S , DAYS

Figure 4
LAPORTE PDU MEOH PRODUCTIVITY VS
CUMULATIVE MEOH PRODUCTIVITY
35

30 i

25 In •

MeOH O
PROD. 20 • •
co
I
co
Co
(G-MOL/KG-H)
15

10 - [] RUN E - 3 ( F 2 1 / O E 7 5 - 3 5 )
• RUN E - 1 ( R 7 1 / O F 1 2 - 2 6 )
_ -"- AUTOCLAVE DATA

0 I I I I I I I
0 50 100 150 200 250 300 350 400
CUM. M e O H PROD., KG M e O H / K G CAT. IN S Y S T E M

Figure 5
LIQUID PHASE M E T H A N O L
1-LITER SLURRY REACTOR SYSTEM
UNBAL,
SYN GAS
r-----"- G
FLoF•w
CONTROL ~"
!
BACK
PRESSURE
REGULATOR
GAS
SAMPLES
GC
PARTIAL

.•ACTIV. REFLUX
(.,,J
I
(.~
BAL. HARCOA
~C, SYN GAS t BEDS
WET
TEST
FLOW METER
:)NTROL
OIL
SAMPLES

/
H 2 RICH L,i
SYN GAS OIL
I FL-OW
CONTROL
1 -LITER
REACTOR
RESERVOIR

A00681.007

Figure 6
EFFECT OF Ni(CO)4 ON METHANOL PRODUCTIVITY

18 R •['-
START OF
Ni(CO) 4 INJECTION

16

MeOH 14 5500 SL/HR-KG


O PROD., 750 PSIG
G-MOL 250°C
12 CO-RICH GAS
"HR-KG

10

8 I I I I I I I I
0 20 40 60 80 100 120 140 160 180
TIME ON SYNGAS, HRS
A00681.009

Figure 7
EFFECT OF F e ( C O ) 5 ON M E T H A N O L PRODUCTIVITY
i

START OF
18 m

•r- Fe(CO) 5
INJECTION
16

MeOH
C,O 14 - 5500 SL/HR-KG
I
PROD. A
750 PSIG
G-MOL
250°C
HR-KG 12--
CO-RICH GAS

10

8 I I I I
0 50 100 150 200 250

A00681,010

Figure 8
EFFECT OF Ni AND Fe POISONING ON METHANOL PRODUCTIVITY

0.98
LAPORTE RUN E-1
0.94

0.90
RELATIVE 0.86 L ~-,~,,,,.~ AUTOCLAVE
CO
I MeOH [] RUNS
PRODUCTIVITY 0.82
[]
I_ [] NICKEL CARBONYL RUN
0.78 • IRON CARBONYL RUN

0.74

0.70
0 200 400 600 800
(Ni+ Fe) ON CATALYST, PPMW
A00682.008

Figure 9
A U T O C L A V E RESULTS WITH A L T E R N A T E LIQUIDS

I
36
IZ: STANDARD PERFORMANCE CURVE,
-r- 32
CO-RICH GAS, 7 5 0 PSIG, 250C
28

0 24
CO
I >-'20
• EXXON H T - 4 3
I-
BB DOW CORNING 2 0 0
>--16
I.- ,~ AMOCO 18-USP
0
~12 C) FREEZENE- 100
0I= 8 A I-'] WITCO- 70
a. /X WITCO LP- 150
-r 4
0
• 0
0 2000 4000 6000 8000 10000 12000 14000 16000
SPACE VELOCITY, S L / H R - K G A00881.013

Figure 10
EFFECT OF A U T O C L A V E STIRRER SPEED ON
M E T H A N O L PRODUCTIVITY
I
20
"I-
,..I
O 15

r.p
I
>: 10
i--
m

m
O LP- 1 5 0 / 4 4 5 0 S L / H R - K G
I--. IN F R E E Z E N E - 100 / 5 5 0 0 S L / H R - K G
A WlTCO-70/4400 SL/HR-KG
E3
O
nr
13. 0
-1- I I
0 200 400 600 800 1000 1200 1400
O
STIRRER SPEED, RPM

Figure 11
EFFECT OF CO 2 ON CATALYST PERFORMANCE
15
CO
CONV.,
%
10

250°C, 750 PSIG


8300 SL/HR-KG
I
35% H2, 51% CO, BALANCE N2
¢:=
O'1
A
W

20
MeOH
PROD.,
G-MOL 15
HR-KG
10 I I ! I I
2 14
INL T C , MOL % A00681,011

Figure 12
METHANOL CO-PRODUCTIONFOR ELECTRIC UTILITY APPLICATIONS

J. F. Weinhold
Tennessee Valley Authority
METHANOL COPRODUCTION FOR ELECTRIC UTILITY APPLICATIONS

J. Frederick Weinhold

The addition of methanol coproduction technology to an integrated coal


gasification combined cycle (IGCC) power plant adds a new dimension to an already
versatile electric generating system. Using available technology, which needs to
be demonstrated at commercial scale using gasified coal, the addition of methanol
coproduction to an economically viable IGCC plant would make sense under favorable
circumstances even at today's depressed oil and natural gas prices. Based on the
Department of Energy's (DOE's) fuel price projections, it would be particularly
attractive in the late 1990s and beyond.

IGCC SYSTEM

The basic IGCC system with advanced gas turbine technology is expected to provide
base-load electric power from coal at efficiencies and costs which are competitive
with conventional pulverized coal with scrubbers, with atmospheric fluidized-bed
combustion systems, and with circulating fluidized-bed combustion systems. In
addition, it offers unique benefits due to its ability to meet very stringent air
emissions standards and to be constructed in a phased manner. Because sulfur
removal is accomplished at pressure under reducing conditions, it is possible to
achieve almost complete capture at reasonable cost using proven chemical process
technology. Thus it would not be necessary to obtain offsets from existing units
when building a new unit under an "umbrella." Nitrogen oxides emissions can be
significantly lower than competing systems through appropriate combustion turbine
design and/or control of combustion conditions.

The phased construction approach allows utilities to schedule the construction of


combustion turbines in response to load growth. Simple cycle gas turbines using

3-49
natural gas can be ordered and installed with as little as a two-year lead time.
Such units are suitable for peaking operations. The utility can increase the
efficiency of the unit at a later date by adding heat recovery steam generators
and turbines. Heat rates of 7000 to 7500 Btu/kWh are possible with natural gas.
This makes the units suitable for intermediate- or base-load operation, provided
that natural gas is available in sufficient quantities. At today's low natural
gas prices this can be the most economic alternative.

Coal gasification can be added to the combined-cycle plant when fuel prices
escalate and system conditions warrant the use of coal. If this capability is
designed in at the start, the gas turbines can provide needed system generation
while the longer lead time gasification and steam units are being constructed.
The result is greater flexibility in responding to changes in demand growth and
fuel availability. Phased construction also allows utilities to minimize rate
shock and spread capital requirements through time. When the time value of money
and inflation are considered, the financial savings associated with phased
construction could account for up to 30 percent of the plant's capital cost.

METHANOL COPRODUCTION

Methanol is currently being produced from natural gas and coal. The raw material
is first converted to a synthesis gas--carbon monoxide and hydrogen--then shifted
to obtain a hydrogen to carbon monoxide ratio slightly over two and finally
recycled through a catalyst bed until it is almost completely converted to
methanol. Modern natural gas-to-methanol plants use nearly two Btu's of feedstock
to obtain one Btu of product.

Coproduction of methanol with electric power offers some real opportunities to


improve efficiency and cut costs. In the once-through concept, the synthesis gas
produced by coal gasification (H/CO = .5) is cleaned and sent through a methanol
catalyst reactor just once. Much of the hydrogen and 20 percent of the total
energy are converted to methanol. The depleted gas is then burned as fuel in the
combustion turbine. The equipment and losses associated with shifting and recycle
are saved.

Methanol coproduction can be included when the gasification unit is designed and
constructed. It can enhance the ability of the unit to meet several utility
objectives.

3-50
Ensure Fuel Availability - Utilities adopting an expansion
strategy involving significantly increased dependance on gas
turbines to meet peaking and intermediate demand face the
possibility that natural gas will not be available at reasonable
prices or at all to meet some system generation requirements.
This may occur soon in response to market disruptions or later due
to the depletion of low cost natural gas reserves. Liquid fuel
must be available to cover this problem. Distillate oil now
provides the alternate, but it is more costly and is subject to
changes in the world oil situation. Even in today's depressed
energy market, methanol from an add-on once-through unit would
provide liquid fuel from coal for peaking gas turbines at costs
which are competitive with distillate. It is possible that the
variable cost of once-through methanol from coal would be
competitive with natural gas. Thus a utility which undertakes an
extensive combustion turbine-based expansion plan would find it
desirable to include once-through methanol capability at its IGCC
plant to supply liquid fuel for its peaking turbines. A once-
through unit associated with a base-load 500-MW IGCC power plant
(normally 400 MW of electricity with the remaining gas used to
produce methanol) would supply approximately 800 to 900 MW of
simple cycle gas turbines operating 500 hours per year for peaking
or 2200 MW if they operated only 200 hours per year. The same
methanol coproduction unit could supply about 250 ~ of advanced
combined cycle generation operating 2500 hours per year for
intermediate load applications. Thus the development of once-
through methanol technology now and making provision for its
inclusion in future IGCC plants allows utilities to build low cost
natural gas turbine peaking plants now and to ensure against
future natural gas/distillate unavailability.

Provide Load Following/Energy Storage - The previous option


assumed that a base loaded IGCC unit would be operated at
essentially full load (85 percent capacity factor) to produce
storable liquid fuel for use in other combustion turbine units in
the system. These other units would operate to meet intermediate
and/or peak demand. It is also possible to design this load
following/energy storage capability into a single IGCC/methanol
coproduction unit. By sizing the combined-cycle plant to handle
the entire output of the gasifiers and providing a bypass of the
methanol unit, it is possible to increase the electrical output of
the IGCC unit by 25 percent while maintaining a constant gasifier
load~ Additional combined cycle or simple cycle combustion
capacity could be included in the unit. The methanol produced
could fuel the combined cycle power unit when the gasifier was not
operating and could fuel additional simple cycle turbines. The
actual matching of gasifier and combustion turbine capacities
would depend on overall system configuration, reliability
requirements, and sparing philosophy. The modular nature of both
the gasifiers and combustion turbines, in conjunction with the
once-through methanol unit, gives the designer a great deal of
flexibility. Thus utilities lacking pumped hydro, compressed air
or other cycling storage potential could meet both base load and
cycling needs with a single IGCC/methanol unit.

3-51
Provide Potential Power C o s t Reductions Through Product
Diversification - Methanol produced at an IGCC/once-through
methanol power plant would be economically competitive with other
sources of methanol, providing that the electric power system was
able to support the cost of the gasification and combined-cycle
units. It is thus feasible for a utility to consider installing a
once-through methanol unit to produce methanol for sale.
Depending on its ultimate use, the methanol may have to be
upgraded to chemical or motor fuel grade. The 500-MW IGCC plant
discussed above would produce about 70-million gallons of methanol
per year, roughly one-third the output of a world-scale natural
gas-to-methanol plant. With today's coal prices, the variable
cost of producing once-through methanol would only be 14 to 25
cents per gallon, well below the current depressed market price of
37 to 42 cents per gallon. This leaves room for a substantial
operating margin even when transportation and upgrading costs are
added. This operating margin could be used to offset some of the
variable costs of producing electricity from the IGCC unit thereby
placing the unit earlier on the dispatch list. If and when
methanol prices increase, the methanol revenue could substantially
reduce the net electric generating cost. Thus the coproduction
and sale of methanol could provide new electric generation from
coal at a net cost approaching the system average cost rather than
well above.

The three options or objectives for methanol coproduction are not meant to be
exclusive. They are aids in thinking about and justifying its installation. It
is quite possible that a once-through methanol unit would be justified in terms of
the insurance it provides a utility system for its natural gas fired combustion
turbines. Once installed, however, it could be used to produce methanol for sale
providing that markets were available, the economics were favorable, and the sales
contracts so structured that it could meet its insurance objective when needed.

ECONOMICS OF METHANOL COPRODUCTION UNDER CURRENT MARKET CONDITIONS

No matter which of the three options or rationales for once-through methanol is


being used, it is necessary to understand the following economic elements.

Variable Costs of Producing Methanol - The variable cost of


producing methanol in a once-through unit can be estimated by
assuming that the system is configured so that the methanol unit
can either be operated or bypassed to produce additional
electricity. The variable cost is then the value of the electric
production foregone plus the variable costs associated with the
methanol unit itself. If the IGCC unit is the marginal producer
in the system at the time, then the value of electricity foregone
is the marginal cost of electricity from the unit. If other more
costly units are operating, then the cost of backing down the IGCC
unit is the marginal cost of the last increment of system supply

3-52
then operating. When the opportunity costs for producing methanol
exceed the system replacement cost, methanol would be produced.
On a system similar to TVA's, located close to eastern coal
fields, coal-based power is generated at a variable cost of 1.3 to
1.8 cents per kilowatt hour (high-sulfur coal available around
$1.25 per million Btu). With highly efficient conversion of
synthesis gas into methanol (efficiency of 95 percent or better),
methanol could be produced at a variable cost of 14 to 19 cents
per gallon ($2.15 to $2.85 per million Btu). On other coal-based
systems with less favorable coal prices, coal-based electric power
is generated at a variable cost of 1.8 to 2.5 cents per
kilowatthour. This would produce methanol with a variable cost of
19 to 25 cents per gallon ($2.85 to $3.85 per million Btu).

Capital Cost of Methanol Unit - The methanol unit itself plus


modifications to the gas cleanup system will require additional
capital expenditures which can be directly associated with the
cost of ~he methanol produced. While there are no definitive
estimates of these costs, they appear to be in the range of 5 to
i0 percent over the IGCC plant cost ($1500/kW). The capital
charge per gallon would depend on the quantities produced per
year. Under favorable methanol demand/market conditions, the
gasifier unit would be run at full capacity to the maximum extent
possible (i.e., 85 percent of the time). The once-through
methanol unit would be run at full capacity (20 percent of the gas
used to produce methanol) except when the utility system
conditions required maximum electric production. At that time,
the methanol unit would be bypassed and full electric production
obtained. Assuming the unit was bypassed 50 times per year for 8
hours each (400 hours), the electric generating unit capacity
factor would be 68.8 percent and 16.2 percent of the rated
gasifier output would be used to make methanol. The use of a 16
percent per year capital charge rate would result in a capital
charge per gallon of 7.7 to 15.5 cents per gallon, depending o n
where within the 5 to i0 percent marginal investment range the
plant was.

Capital Cost of Gasification and Power Generating Units - During


normal once-through methanol operation with the gasifiers
operating at i00 percent of capacity, the system would use up to
20 percent of the gas stream energy for methanol production and 80
percent for electricity production. It is thus necessary to
decide how much, if any, of the capital cost of the gasification
and related units should be assigned to the methanol produced.
The capital costs of power production and gasification units are
roughly the same, so the capital cost of 20 percent of the
gasification unit would be about I0 percent of the total plant
cost. This is roughly equal to the cost of the methanol
coproduction unit. The once-through methanol unit would, however,
be operated in such a way as to increase the overall plant
capacity factor. Given adequate markets for the methanol and the
favorable marginal cost/value picture, it would be possible to
increase the capacity factor of the IGCC system to about 85
percent, or what a chemical plant might achieve, as compared with
a 60- to 70-percent rate for utility generating units. (Eighty
percent of 85 percent is 68 percent.) This could justify power
production bearing the entire capital cost of the gasification
unit. Such an allocation would be justified where the combustion

3-53
turbines are designed to handle the full flow of the gasifiers
(methanol unit bypassed). This configuration could be viewed as a
normal IGCC system with an add-on once-through methanol unit to be
operated when the system did not need the full power output. The
system would be operated for more hours per year at 20 percent
lower electrical output to produce about the same number of
kilowatt hours per year.

Value of the Methanol - The methanol could be used directly by the


utility in combustion turbines to replace distillate or natural
gas making it a coal-derived fuel suitable for peaking service.
If the capital charges associated with modifying the fuel storage,
fuel handling, and burner configuration were neglected, then the
Btu value of the methanol would equal the Btu value of the
distillate or natural gas. From 1983 through 1985, distillate
sold for 75 to 85 cents per gallon ($5.40 to $6.10 per million
Btu) which is equivalent to 35 to 40 cents per gallon of
methanol. The average cost of natural gas to utilities was $3.50
to $3.75 per million, but varied with location and type of
contract. This is the equivalent of 22 to 24 cents per gallon of
methanol. The world oil market is currently experiencing a rapid
price drop, from nearly $30 per barrel to less than $20. This
could result in distillate prices around 50 cents per gallon.

The current methanol market price is 37 to 42 cents per gallon


($5.75 to $6.50 per million Btu). While the stabilized methanol
fuel produced in a utility once-through unit would be suitable for
use in combustion turbines with no further processing, it would
have to be upgraded, i.e., distilled, for sale as chemical-grade
methanol. This would require additional capital and add about 2
cents per gallon to operating costs. It is not clear how much
upgrading would be required for the methanol to be used in
gasoline blending. To sell methanol, it would also be necessary
to incur marketing and transportation expenses. It would cost
about 5 to I0 cents per gallon to barge methanol from accessible
inland locations in the eastern U.S. to Gulf Coast markets.
Special conditions of convenience, such as nearby users, could of
course offset this cost. The marketing costs would involve
storage, handling, and administrative costs associated with
particular customers or markets.

The value of coproduced methanol fuel is currently below the level


it would have been in 1980 through 1982. Under the economics
which prevailed from 1983 through 1985, its value would be 35 to
40 cents per gallon ($5.40 to $6.10 per million Btu) as a
replacement for turbine distillate (2.15 gallons of methanol to 1
gallon of distillate) and 26 to 35 cents per gallon (38 to 42
cents - 5 to I0 cents transportation - 2 cents upgrading) as
merchant-grade methanol ($4.00 to $5.40 per million Btu). These
figures do not reflect the added capital needed to use or market
the methanol. On this basis, internal use would be slightly more
attractive than sale. However, if distillate prices remain at
their current (early 1986) low levels and methanol continues to
retain its market price, then sale would be more attractive.
Given this uncertainty, both options should be evaluated.

Capital Cost of Using/Selling Methanol - The use of methanol in


the utility's own combustion turbines could involve modifications

3-54
p
to fuel storage, fuel handling, and burner systems, since it has
about half the energy per unit volume as distillate fuel oil. If
many existing units must be modified, the investment would be
substantial. General Electric has estimated the cost to be about
$i.5 million for modifying one of their existing large combustion
turbines. The methanol could also be used for lighting off and
sustaining coal-fired units. The modification and capital costs
for this application have not yet been estimated.

The investments required to upgrade, handle, and sell methanol


have not been estimated either. Generally, the unit costs are

i acceptable where large quantities and high capacity factors are


involved.
Other Economic Factors To Be Evaluated - A number of secondary
concerns can also affect the overall economic picture. Since the
gasification units and the combustion turbines are affected by
ambient temperature differently, stored methanol could offset the
mismatches. An IGCC unit designed to produce sufficient gas to
fully load the combustion turbine/combined cycle in the winter
(20°F) with the methanol unit bypassed would have sufficient spare
gas production capacity in the summer (90=F) to fully load the
turbines (reduced output) with the methanol unit in full
operation. Supplemental firing of stored methanol to produce
steam for the steam turbine generator could also increase unit
output.
Likewise, methanol coproduction could be used to eliminate the
need for spare gasification capacity while assuring a high degree
of plant availability. The power unit could be operated with
stored methanol when one or more gasifiers are out of service in
order to maintain output. The methanol could also be used for
supplemental firing if a turbine were out of service.
Furthermore, with one of five normally operating gasifiers down,
the methanol unit could be bypassed and the remaining gasifiers
used to meet i00 percent of the electrical load.

Finally, economies of scale in building the gasification and/or


power units of the plant may have a differential impact on the
different configurations and options.

Summary - The table below summarizes some of the costs and values
of coproduced methanol under current market conditions, assuming
that the methanol does not have to bear a share of the capital
cost of the coal gasification unit.

I
i

3-55
COAL COSTS

Favorable Less Favorable


Methanol Costs Cents per Gallon

Variable Costs 14-19 19-25


Total cost at 5% added capital 22-26 26-33
10% 29-34 34-40

Methanol Values

As distillate substitute @ 75-85 cents/gal. 35-40


@ 50 cents/ gal. 23
As natural gas substitute 19-23
To be upgraded and sold as
chemical methanol 25-35

ECONOMICS OF METHANOL COPRODUCTION IN THE FUTURE

The previous discussion focused on the economic competitiveness of methanol


coproduction in face of today's depressed oil and gas prices. Under especially
favorable circumstances it would be economically competitive. However, the real
justification for developing synthetic fuels from coal rests on the widely held
e~pectation that oil and gas prices will again rise significantly faster than
inflation due to resource depletion. Coal prices, on the other hand, are expected
to remain stable due to the vastly larger coal resource base. DOE shows such a
change taking place in the 1990s, as evidenced by the reference fuel price data
imcluded in the Clean Coal Technologies Solicitation. This data (in constant 1984
dollars) indicates that oil prices will fall from their 1984 levels to 1990 and then
rise rapidly through 2010. Natural gas prices rise slowly until 1990 and then rise
rapidly with oil prices. Coal prices, however, rise only moderately during the
entire period.

These fuel prices can be used to estimate coproduced methanol costs and values by
employing techniques similar to those employed in the previous analysis. Assuming
that the delivered cost of coal to an IGCC/once-through methanol unit ranges
between ll0 percent and 150 percent of the average minemouth cost and that
incremental capital costs are between 5 and 10 percent of the cost of an IGCC
plant, the range of total costs is 25 to 37 cents per gallon in 1984. Based on
the DOE data, these costs increase at a rate of less that 1 percent per year.

3-56

p kl ¸
Beginning in 1990, the value of methanol as a distillate substitute or in
competition from methanol from natural gas increases at nearly 5 percent per
year. As shown in figure i, these conditions make coproduced methanol very
attractive in the post 1995 period. By 1995 under virtually all conditions, it is
less costly than methanol from natural gas or distillate. By 2000 it can replace
natural gas as a combustion turbine fuel on a full cost recovery basis.

CONCLUSION

Methanol coproduction adds significant versatility to an already versatile IGCC


power generating system using low-cost, high-sulfur coal. The most fundamental
questions utility planners face involve the viability of the basic coal
gasification unit and the potential uses of methanol. When does an IGCC c o m p e t e
with natural gas fired turbines and other new coal based generating options in a
particular system? This involves environmental considerations and other
imponderables. Phased generation additions, starting with natural gas fired
combustion turbines followed by steam bottoming cycles and coal gasifiers when
fuel availability and economics justify them, may be the prudent answer.

Once this first hurdle is passed, does the utility have need for a storable liquid
turbine fuel or can it find good markets in which to sell the methanol? A 500-MW
IGCC plant with a once-through methanol unit operated in the manner described
above, would produce 70 million gallons of methanol per year, one-third the output
of a world-scale methanol from natural gas plant. If the utility has acccess to
relatively low-priced coal as well as methanol uses and markets which support
methanol values comparable with distillate or merchant methanol, then methanol
coproduction would be a good investment.

Even at today's depressed oil and natural gas prices, the addition of once-through
methanol capability could be justified under favorable circumstances including an
internal need for the methanol to replace distillate or a market with prices
linked to those for chemical-grade methanol. Based on the DOE fuel price
projections, coproduced methanol would be competitive under almost all
circumstances by the late 1990s. This is the time methanol units associated with
phased construction IGCC plants would first be expected to come on line. Once
installed, these units would produce methanol at the lowest variable cost of any
domestic source, so would be operated at full load except when system electric
power Feeds dictated that high cost peaking power was required.

3-57
90

PROJECTED METHANOL
PRODUCTION COST AND VALUES
80 (BASED ON DOE F U E L PRICE PROJECTIONS)

Z- 70
©
_.t
J
<
L9
Pr" REPLACEMENT
uJ VALUE
c,3
60
k.-
z
LU
L,)
V A L U E IN
00
~3b METHANOL VALUE
50 MARKET , REPLACEMENT j
LU
NATURAL /
J
<
>
(.,o
O 40
L)

"F
cu

30

COST OF
COPRODUCED
METHANOL
20

10

1985 1990 1995 2000 2005 2010


YEAR

3-58
It is necessary to optimize the IGCC/methanol unit designs through continued
design efforts and then through a large-scale demonstration such as has been
proposed at TVA's Muscle Shoals gasification facility. The testing would also
verify the operation of the gas clean up and methanol units under utility
conditions (i.e., transient or bypass operation). The design and experimental
data would permit utility system simulations and sensitivity studies to assess the
merits of the fuel substitution, methanol sales, and energy storage options in
specific systems.

Transcending the longer term economic benefits of specific applications of once-


through methanol is the short term need that utilities committed to building
natural gas fired gas turbines to meet peak loads have for a credible coal-based
fuel option if and when natural gas becomes unavailable. The results of the
current and proposed development programs, combined with provisions for adding
coal gasification and once-through methanol production to future gas turbine
combined-cycle installation, will give them the assurance they need to chose the
low cost solution today without the risk of future fuel unavailability.

SUMMARY OF ECONOMIC ANALYSES


CURRENT FaiRKET CONDITIONS

Fuel Production

Medium Lo__ww
Methanol Costs (~n cents)

Electric power cost per kWh 2.5 1.8 1.3

M e t h a n o l p e r gal at 95% Conversion* (9.16) 22.9 16.5 11.9

Add variable operating costs


2 cents/gal. 24.9 18.5 13.9

Equivalent distillate (cent/gal) (2.15


Gal. MEOII/Cal. distillate) 53.5 39.8 29.9

Equivalent natural gas ($/106 Btu) 3.85 2.85 2.15


(Also equivalent distillate cost
i n $/106 Btu.)

* Based on advanced cycle heat rate for combined cycle of 7420 Btu/kWhr.

3-59
SUMMARY OF ECONOMIC ANALYSES
CURRENT MARKET CONDITIONS
(Continued)

Capital Costs

Assumptions

IGCC plant cost 15001kW


Annual capital charge rate 16%
Kilowatthours forsaken to produce 1 gallon
of methanol (@ 95% production efficiency) 9.16

Peaking Applications 50 times/year 08 hours = 400 hrs


or 4.5% of time
Casification Plant Capacity Factor 85%
Once-Through Methanol Production 20%

Results

Methanol Production 85% x 20% x 95.5% = 16.2%


16.2% x 8760/9.16 = 155 Gal/yr/kW

Electric Production
85% (95.5% x 80% ÷ 4.5% x 100%) 68.8%

Capital Charge 16% x 1500 $240/year


at 10%: 240 x 10%/146 15.5 Cents/gal
5%: 240 x 5%/146 7.7 Cents/gal

EnerRv Storage

Assumed Unit Performance: Efficiency Heat Rate

Coal Gasification + AGR 76.9%


Once-Through Methanol Unit 95+%
Advanced Combined Cycle 46% 7,420
Simple Cycle Gas Turbine - Natural Gas 28.7% 11,900
Methanol
- 29.6% 11,516
Methanol HHV in Gas Turbine
versus ~ G 103%

Resulting System Performance:


IGCC: bypass MEOH unit 36.7% 9,300
ICCC with OTM in operation 36.7%
Methanol through combined cycle 33.6% 10,150
Resulting turnaround efficiency 91.6%
Methanol through simple cycle 21.6% 15,780
Resulting turnaround efficiency 58.9%

3-60
PROJECTED METHANOL COSTS AND VALUES
BASED ON DOE REFERENCE DATA)
(1984 Dollars)

PRODUCTION COST 1984 1990 1995 2000 2005 2010

Minemouth Coal Price ($/106 Btu) 1.18 I .28 1.31 1.43 I .49 i .56

Variable Cost of MEOH 1 (cents/gal.)


@ +10% transportation .158 .167 .174 .181 . ! 87 . 194
@ +50% transportation • 198 .211 .222 .230 .238 .247
Total Cost
@ 10% transportation 5% capital .235 .244 .251 .258 .264 •271
@ 50% transportation 10% capital .353 .366 .375 .385 .393 .402

VALUE
Wellhead Gas Price ($/106 Btu) 2.60 2.76 3.68 4.80 5.70 7.68

Delivered To Gulf Coast Chemical 3.10 3.26 4.18 5.30 6.30 8.18
Oo
I Plant (+$.50/106 Btu)
C%

Cash Cost of Chemical Methanol .42 .43 .53 .65 .75 .95
from Gas 2 ($/gal)

Value of Methanol Sold 3 ($/gal)


@ .05 cents/gal Barge .30 .31 .41 .52 .62 • 82
@ .10 cent/gal barge .35 .36 .46 .57 .67 .87

Gas to Utilities (+ $1/106 Btu) 3.60 3.76 4.68 5.80 6.70 7.68

Value as Natural Gas Turbine Fuel .231 .243 .302 •375 .433 •561
Replacement ($/gal)

Distillate to Industry ($/gal) .86 •69 •90 1 •10 1 •40 1.69

Value as Distillate Turbine Fuel .400 .321 .419 .512 .651 .7 86


Replacement ($/gal)

Notes: I. ([coal cost x .0093 + .003] x 9.16 + •02)


2. (.1055 NG cost + •09)
3. ( -.02 upgrading - transport)
IGCC CRITERIA FOR ONCE-THROUGHMETHANOLUSING THE LPMEOHPROCESS

R. L. Mednick

Chem Systems

and

T. L. Wright, H. M. Weatherington and J. Pech


Tennessee Valley Authority
ABSTRACT

Integrated Gasification Combined Cycle (IGCC) f a c i l i t i e s such as the Cool Water


plant have proven to be a clean, efficient, and economic means of generating
electric power from coal. The IGCC f a c i l i t y has environmental advantages over
conventional pulverized coal or fluidized bed combustion especially with the more
stringent air pollution controls now being contemplated because of acid rain
concerns. Although IGCC f a c i l i t i e s would be roughly equivalent in capital
investment to conventional coal-fired steam plants, the resulting cost of power
would be lower due to higher efficiency. Flexibility in the IGCC f a c i l i t y ,
however, may be somewhat more expensive. One efficient way to provide f l e x i b i l i t y
is to convert some of the energy from the gasifier into a storable liquid such as
methanol. The once-through methanol (OTM) concept is being developed with this
application in mind.

This paper discusses some of the key design issues in integrating an OTM unit into
an IGCC plant. Also, plans for demonstrating OTM synthesis on actual coal-derived
synthesis gas at TVA are outlined.

3-65
I. INTRODUCTION

IGCC/OTM DESCRIPTION

Future electrical generating plants will include IGCC plants to provide base load
electric power from coal. IGCCoffers unique benefits to meet very stringent air
emission standards especially for SO2. An IGCC plant can be constructed in phases
allowing u t i l i t i e s to install natural gas fed combustion turbines for peaking
operations and later add heat recovery steam generators, steam turbines, and coal
gasification/gas cleanup units. This approach enables u t i l i t i e s to spread capital
investment over time and delay capital investment decisions. Installation of OTM
production into an IGCC plant allows partial conversion of the fuel gas to
methanol. The unconverted fuel gas from the OTM process can be combusted in the
combined cycle plant. I t is well established that methanol provides a clean,
economic, and secure source of liquid fuel that can be stored and then later
burned in a combustion turbine to produce electrical power for peaking or load
following. The crude methanol produced can also be upgraded and sold as
chemical grade methanol.

Figure I shows a general diagram of an IGCC/OTM plant. The IGCC plant is composed
of a gasifier and its waste heat recovery (WHR) unit, and acid gas removal (AGR)
unit, and a combustion turbine with a WHR system. The IGCC could be modified by
adding an OTM process after the AGR system to prevent the methanol catalyst from
being poisoned by sulfur compounds. The OTM unit would be composed of a guard bed
system, a methanol synthesis and recovery section, a methanol storage area, and
could include a peaking combustion turbine. In the methanol reactor, two moles of
H2 and one mole of CO react over the copper-based catalyst to form one mole of
methanol (CH30H) which is condensed as a l i q u i d . Up to 25 percent of the energy
in the fuel gas can be converted to methanol. The unconverted fuel gas (depleted
synthesis gas) can be saturated with water vapor and then burned in the combustion
tuFbine of the conventional combined cycle plant.

3-67
(~POWER

STEAM J STEAML

STEAM STEAM
COAL 02
BYPASS

w.s.. H c,o o.c


~,wI~
COMBUSTIONL~
HEAT GAS THROUGH WHA
ES
ATE F~LUE
CATION RECOVERY REMOVAL METHANOL TURBINE I I RECOVERY
I i GAS
CO

SLAG SULFUR UNCONVERTED


FUEL GAS (~1~ POWER

¥
METHANOLi__1~COMBUSTION~
POWER
STORAGE I I TURBINE I "~"
(PEAKING)

FIGURE I
IGCC / OTM SCHEMATIC
Conventional methanol synthesis requires shifting a portion of the CO to H2
forming a chemically balanced gas (two moles H2 to one mole CO produces one mole
of CH3OH.) It also recycles the unconverted synthesis gas at a high ratio to feed
gas to maximize methanol production, since only partial methanol conversion occurs
during each methanol reactor pass. Thus, recycling the unconverted synthesis gas
allows almost complete conversion of the H2 and CO to methanol in a conventional
plant.

OTM synthesis d i f f e r s from conventional methanol production since the synthesis


gas is not shifted. This results in a CO-rich gas that is p a r t i a l l y converted to
methanol in a single pass with the unconverted fuel gas available for IGCC power
production. A schematic showing conventional versus OTM methanol production is
shown in Figure 2.

LPMEOH PROCESSDESCRIPTION

Gas phase processes are used for conventional methanol production. These
processes use various means of temperature control, including a recycle gas stream
as a dilutent to control the exothermic heat of reaction. With the OTM concept
there is generally no recycle gas to moderate the heat release. Since methanol
conversion decreases as temperature increases and since state of the art catalysts
are very temperature sensitive, the a b i l i t y to control temperature in an OTM
operation is very c r i t i c a l . The Liquid Phase Methanol (LPMEOH) process maintains
excellent temperature control by suspending the methanol catalyst in an inert
hydrocarbon liquid which acts as a heat sink. Figure 3 shows the LPMEOHreactor
configuration which was invented by Chem Systems and operated by Air Products at
the LaPorLe Process Development Unit (PDU). The catalyst is entrained in the
inert liquid as a slurry, recirculated from the reactor to a waste heat boiler for
steam production, and returned to the reactor. The 482OF fuel gas from the
reactor is cooled to condense vaporized hydrocarbon and methanol. The
feed/pro(Juct gas heat exchanger preheats the fuel gas from AGR prior to the LPMEOH
reactor and the final gas exchanger reheats the fuel gas for use in the combined
cycle. The crude methanol produced contains at least 94 percent methanol, less
than two percent higher alcohols, up to four percent water, and a trace amount of
inel't hydrocarbon liquid. The condensed hydrocarbon liquid is separated from the
crude methanol and recycled back to the LPMEOHreactor. A trace amount of
hydrocarbon remains in the crude methanol after separation, but this small amount
may prove to be beneficial as a lubricant for the combustion turbine.

3-69
STEAM WPRESSOR

GkSIFIER
tH H H
WHR
CO
SHIFT
AGR PURGE

METHANOLF CHEMICALGRADEMETHANOL
DISTILLATION

!
,,,,,,,.j
STEAM
0

GASIFIER WHR COMBINED


AGR THROUGH "--~ATMOSPHERE
METHANOL CYCLE

®
CRUOE
METHANOL COMBUSTION
STORAGE TURBINE

FIGURE 2
CONVENTIONAL VERSUS ONCE-THROUGH METHANOL PRODUCTION
OESULFURIZED
SYNTHESIS G~,S
O ~- FINAL GAS
EX ~-~
DEPLETED
FEED/PRODUCTGAS SYN GAS

C.

STEAM T
LPMEOH ~ )
REACTOR HB V/L
I I OIL SEPARATOR
(.~
I

OIL RETURN FUEL GRADE


PUMP METHANOL
PRODUCT

CONDENSED
SLURRY OIL PUMP
PUMP

FIGURE 3
THE LPMEOH PROCESS WITH A
LIQUID- ENTRAINED REACTOR
2. IGCC/OTM APPLICATIONS

The crude methanol produced in an OTM unit provides a reliable and secure source
of a clean, storable fuel from coal. As a storable energy form, methanol can be
used in combustion turbines for peak and intermediate demand loads and for load
following. The crude methanol also offers d i v e r s i f i c a t i o n opportunities. With
upgrading, the methanol can be sold as chemical grade or fuel grade methanol.

Considering today's rapid changes in energy pricing, the r e l i a b i l i t y of fuel


supply is c r i t i c a l . IGCC/OTMoffers many fuel options to e l e c t r i c u t i l i t i e s even
at this time. With "cheap" and abundant natural gas, peaking combustion turbines
can be installed to meet short-term load growth requirements. When natural gas
prices increase, the a v a i l a b i l i t y w i l l decrease, and combustion turbines can then
be modified to a combined cycle plant by the addition of waste heat recovery. As
the energy pricing structure in the 1990s returns to i t s anticipated higher
levels, a gasification/gas cleanup plant can be added to the combined cycle plant
making an IGCC plant. An OTM unit can then be added to the IGCC plant to provide
a secure source of storable fuel.

3-73
!

..... !
3. IGCC/OTM DESIGN CRITERIA

For an IGCC/OTM plant, the design c r i t e r i a may d i f f e r from a "standard" IGCC plant
design. The proposed OTM test program at TVA (see Section 4) w i l l provide data to
establish many of the design c r i t e r i a .

All the technical data and evaluation in t h i s section were obtained from published
reports, technical a r t i c l e s , and vendor information.

GASIFICATION

The selection of a g a s i f i c a t i o n process w i l l have an impact on the design of an


OTM unit in an IGCC plant. The most important parameters are the fuel gas
pressure, H2 to CO r a t i o , and CO2 content. The basic design difference between
the available g a s i f i c a t i o n technologies is the choice of wet or dry coal feed. A
wet coal feed process, such as Texaco, w i l l produce a much higher CO2 content in
tile fuel gas compared to a dry coal feed process. Based on TVA's experience with
a Texaco g a s i f i e r using bituminous coals, the fuel gas CO2 content can range from
14 to 22 percent depending on the specific coal and the g a s i f i e r operating
temperatuFe. A dry coal feed g a s i f i c a t i o n process, such as Koppers-Totzek
(atmospheric or pressurized) or Shell (pressurized), produces a fuel gas with a
CO2 content of 2 to 8 percent ( ! ) .

Fixed-bed g a s i f i c a t i o n , such as Lurgi, produces a fuel gas with a CO2 content of


30 or 4 percent, depending on whether i t is dry bottom or slagging Lurgi,
respectively ( I ) . The effect of CO2 content on OTM conversion with CO-rich gas
has not been f u l l y determined at t h i s time but is c u r r e n t l y being investigated.
However, for conventional methanol production on a balanced gas, a minimum of 3
peFcent CO2 is Fequired.

3-75
The different gasifiers w i l l also produce synthesis gas with a range of H2 to CO
ratios. As the H2 to CO ratio increases, the OTM conversion efficiency w i l l
increase. The CO-rich synthesis gas produced by a Texaco g a s i f i e r has
approximately the same H2 to CO ratio as almost all other advanced gasifiers.

Gasifiers for the chemical process industry operate up to 1000 psig. For methanol
synthesis, the equilibrium is favored at higher pressures (750 - 1500 psig) rather
than the lower pressures required for combined cycle operation (2). The higher
pressure of the IGCC/OTM f a c i l i t i e s would also improve the AGR efficiency. The
combustion turbine part of an IGCC plant requires fuel gas at 300 psig for optimum
efficiency. In all likelihood, economic studies may show that the higher
pressures are required for OTM synthesis. Thus, an expander may be needed to
produce power as the gas pressure is reduced from the OTM unit to the combined
cycle plant.

Gasification produces two forms of sulfur compounds: hydrogen sulfide (H2S) and
carbonyl sulfide (COS). Usually 95 percent of the reduced sulfur is H2S. For the
NSPS compliance for SO2, only the removal of H2S is required. However, for OTM
operation removal of both H2S and COS is required to prevent catalyst damage. For
AGR systems which do not e f f e c t i v e l y remove COS, COS conversion to H2S is required
prior to AGR. Existing gasification processes use a wet scrubber as part of their
particulate removal system. From the wet scrubber, preheat of the COS hydrolysis
catalyst above the water dewpoint is required to prevent catalyst damage.
Alternately, COS could be hydrogenated and removed as H2S in a guard bed upstream
of the OTM unit. Consequently, the IGCC WHR system design should be integrated
with the OTM design requirements. For instance, in a quench Texaco gasifier,
sufficient heat is not available for COS hydrolysis preheat. However, in a Texaco
gasifier with radiant/convective heat recovery, s u f f i c i e n t heat is available for
COS hydrolysis preheat.

ACID GAS REMOVAL

The selection of an acid gas removal (AGR) system for a commercial IGCC plant
should consider sulfur (H2S and COS) and trace methanol catalyst poison removal,
H2S versus CO2 s e l e c t i v i t y , and proven commercial experience. Data are not
available from coal-derived synthesis gas to determine the trace methanol catalyst
poison removal capability of all the various AGR processes.

3-76
Physical absorption processes appear to be the best suited for the IGCC operating
conditions and sulfur removal requirements. Of the physical absorption processes,
only Rectisol, which uses cold methanol absorbent, and Selexol, which uses
dimethylether of polyethylene glycol, have been proven with coal-derived gas.

Rectisol I f , using a mixture of methanol and toluene, may be more selective to H2S
versus CO2 than Rectisol, but i t has not been proven commercially (~). Purisol,
physical absorption with N-methyl-2-pyrrolidone (NMP), is very selective toward
I12S over CO2 but does not readily absorb COS. Although i t has been used
commercially, Purisol has not been used on coal-derived gas. Activated Purisol
uses an activator to catalyze the reaction of COS to H2S at a higher solvent
temperature (4). Sepasolv, which uses a mixture of oligoethylene glycol methyl
isopropylethers, is very similar in chemical structure and physical properties to
Selexol solvent. COS and mercaptans are more soluble in Sepasolv than in Selexol.
H2S s o l u b i l i t y is higher for Sepasolv, which allows greater CO2 slippage for
equivalent solvent flowrates. Like Purisol, Sepasolv has also been used
commercially, but not with coal-derived gas (5).

Figure 4 shows a comparison based on an internal TVA study for EPRI of the various
AGR systems for H2S versus CO2 s e l e c t i v i t y . Several examples are shown for
Purisol, Rectisol, and Selexol indicating the effect of solvent flow and
temperature (Purisol and Rectisol), and AGR configuration (Selexol). Studies are
required to compare the costs of greater sulfur removal in the AGR unit versus the
size and change-out period for the s a c r i f i c i a l guard beds. Studies are also
required to assess the cost of H2S versus CO2 s e l e c t i v i t y in AGR systems on both
gas turbine power generation and H2S acid gas concentration in the sulfur recovery
unit feed gas. In a conventional methanol plant, CO2 content of the synthesis gas
should be about three percent. The AGR systems are designed to remove both
H2S/COS and CO2. In an IGCC/OTM plant, CO2 removal should be minimized since the
additional mass provided by the CO2 in the fuel gas provides increased power
output in the combined cycle plant. However, increased slippage of CO2 through
the AGR system usually corresponds to less sulfur removal. Consequently, the
selection and design of the AGR system must maximize the removal of H2S/COS and
minimize the removal of CO2. This AGR design would also be advantageous to any
sulfur recovery unit since a concentrated H2S feed would be produced.

With high CO2 slippage, removal of sulfur compounds in AGR systems below 0.1 ppm
or less are generally not obtained. Guard beds of materials such as zinc oxide
are required to obtain sulfur removal to lower levels. Since guard bed materials
have to be changed after becoming loaded with sulfur, there is a trade-off between
the AGR sulfur removal efficiency and the guard bed size/change-out rate.

3-77
7(3

60- RECTISOL

, LI- 501
Z

~ 4I...-0
ACTIVATED
-J Z
0 -- PURISOI.(~
W SELEXOL ~ ,A
m 30 PURISOL•
0 (CASE I) R~EC'TISOL II-
I
rn SELEXOL
(CASE 2 A ) v
N
0
u 20
• PURISOL SEPASOLV M P E ~

I0 SELEXOL
(CASE 2B)v

0 I I I I I
6 5 4. 3 2 I 0

PPMv TOTAL SULFUR IN PRODUCT GAS

FIGURE 4

COABSORPTION OF CARBON DIOXIDE VERSUS TOTAL SULFUR REMOVAL


PROTECTIVE GUARD BEDS

Methanol synthesis catalyst can be rapidly deactivated by certain trace components


found in coal-derived synthesis gas, even after AGR. Table 1 shows the allowable
impurity levels in the feed gas. The limit of 0.06 ppmv for total reduced sulfur
necessitates the use of an AGR unit followed by a sacrificial guard bed such a
zinc oxide. The low allowable limit for nickel and iron carbonyls also requires a
guard bed and special consideration for materials for construction. Halides and
HCN can be reduced to acceptable levels in an alumina or copper catalyst guard
bed.

Table I
ALLOWABLE IMPURITIES IN METHANOLFEEDGAS

Trace Component Design Limits


Required for Methanol Synthesis (DDmv)
United Chem Systems/
Component Catalysta APCI LaPorte
Total sulfur 0.06 0.06
(H2S + COS)
Halides 0.01 0.01
Cl as HCI
Unsaturated 300.0 -
Hydrocarbons
Acetylene 5.0 5.0
NH3 10.0 I0.0
NOx 0.I 0.i
HCN 0.01 0.01
Fe, as Carbonyls 0.04 0.01
Nickel, as Carbonyls 0.01 0.01

aFor conventional gas phase methanol plants.

The choice of guard beds used in a commercial plant depends on the removal of
methanol catalyst poisons by the AGR unit. A low temperature zinc oxide guard bed
should effectively remove H2S at operating temperatures as low as 400OF. Small
quantities of COS can effectively be removed at 400°F by a hydrolysis

3-79
reaction to H2S. However, i f a substantial quantity of COS is present, the high
temperature zinc oxide bed may be required with operation at about 700OF (6,7).
One alternative to a high temperature zinc oxide bed for COS removal is a
three-layer guard bed composed of zinc oxide for H2S removal, cobalt molybdenum
catalyst to hydrogenate the COS, and zinc oxide for residual H2S removal. A
second guard bed may be required i f iron and nickel carbonyls are present.
Stainless steel materials of construction would be necessary in areas that are in
the critical temperature range for carbonyl formation.

A l i s t of various trace contaminants obtained during TVA's operation at Muscle


Shoals are presented in another paper at this conference entitled "TVA Gas
Processing Measurements In Support of Methanol Production From Coal."

METHANOL PRODUCTION

A conventional methanol unit uses gas recycle to control temperature and to


prevent catalyst deactivation. In an IGCC/OTM unit, methanol production would
occur during a single pass with no gas recycle. Without gas recycle in a
conventional methanol unit, heat removal duty and temperature gradients around the
catalyst w i l l be more severe.

in the conventional ICI methanol reactor (Figure 5), the gas temperature rises
from the heat of reaction across the top catalyst bed. Cooled feed and recycle
gas are injected between the f i r s t and second beds to reduce the temperature and
consequently restore optimum methanol conversion and prevent catalyst damage.
This procedure is repeated throughout the length of the reactor. A temperature
gradient exists through the depth of each bed (8).

In the conventional Lurgi boiling water reactor, better isothermal conditions are
maintained than in an ICI reactor. The catalysts are inserted into vertical tubes
which are surrounded by boiling water. The temperature of the synthesis gas
passing through the tubes is controlled by the temperature and pressure of the
steam generated from the boiling water. A temperature gradient exists across the
tube diameter (9).

The LPMEOHprocess offers the potential advantage of better heat removal compared
to conventional gas phase methanol processes. In the LPMEOHreactor, inert
hydrocarbon liquid surrounds each suspended catalyst particle and removes the heat
of reaction e f f i c i e n t l y while maintaining an optimum catalyst temperature. The
LPMEOH process is also not affected by low gas flow during OTM turndown since the

3-80
CONVENTIONAL GAS PHASE REACTORS:

ICI RG__.._~I
LU

SYNTHESIS GAS FEED AND RECYCLE STEAM

TO HEAT8
BFW ~'-~PRODUCT
RECOVERY
TO HEAT
a PRODUCT SYNTHESIS GAS FEED
RECOVERY
AND RECYCLE

LPMEOH REACTOR:

STEAM

~ S Y N T H E S I S GAS FEED

TO HEAT &
PRODUCT
RECOVERY

FIGURE 5
METHANOL REACTOR SCHEMATICS

3-81

. . . . . . . . . |
slurry in circulation remains turbulent. This provides efficient heat transfer
during turndown, preventing catalyst damage. Another potential advantage of the
LPMEOH process is the use of continuous catalyst addition and withdrawal to
maintain a high level of catalyst productivity without shutdown for catalyst
replacement.

In a conventional gas phase methanol process the catalyst is replaced at periodic


intervals, every three or four years. However, methanol production decreases
sharply after i n i t i a l operation and continues decreasing throughout the catalyst
life,

Integration of heat recovery is the key to efficient placement of an OTM unit in


an IGCC plant. The medium pressure steam produced by the methanol reaction is
available for use by the IGCC plant for power generation or fuel gas reheating and
saturation (see Figure I). Since the unconverted fuel gas from the OTM unit has
been cooled to condense the methanol, i t must be reheated before entering the
expansion or combustion turbine. I f the OTM unit is operated at 750-1500 psig,
then an expansion turbine would probably be'required and the gas entering the
expansion turbine from the methanol unit would require heat. Consequently, OTM
operation at lower pressures may not economically j u s t i f y an expansion turbine.

Crude methanol production from conventional plants using a balanced synthesis gas
contains less water and less higher alcohols compared to an OTM using CO-rich gas.
In an OTM plant with a higher CO2 content in the synthesis gas feed, the crude
methanol produced contains more water because of the secondary reaction of H2 and
CO2.

3-82
4. LPMEOHDEMONSTRATIONAT TVA

TVA ADVANCEDCOAL GASIFICATION TECHNOLOGY (ACGT) FACILITY

The TVA Muscle Shoals' gasification f a c i l i t y was originally designed to determine


the technical, economic, and environmental aspects of substituting coal for
natural gas as feedstock for manufacturing ammonia. The coal gasification
Facility gasifies 200 tons of coal per day and produces 10 million standard cubic
feet per day of carbon monoxide and hydrogen. The plant uses the Texaco coal
gasification p~'ocess and is sufficiently flexible to test bituminous coals with
different heat, ash, and sulfur contents and with different grinding
characteristics. The ACGT f a c i l i t y under the Ammonia from Coal Project has
operated 3,780 hours during 90 test periods. Figure 6 is a process flow diagram.

The 60-plus percent coal-water slurry is pumped to the gasifier where i t reacts
with oxygen to produce a synthesis gas of H2 and CO at about 2500OF. The gas is
scrubbed in a water separator for particulate removal. Sulfur in the coal forms
reduced sulfur compounds, H2S and COS. For ammonia production, hydrogen is
required and shift reactors convert CO to H2. For the OTM project, the shift
converters wi]] be bypassed to produce a CO-rich fuel gas that is required in an
IGCC plant.

A COS hydrolysis reactor converts almost all of the COS in the synthesis gas to
H2S prior to entering the AGR unit. A Selexol AGR system reduces the synthesis
gas sulfur level to less than I ppmv. Less CO2 w i l l be produced since the
CO-shift reactors w i l l be bypassed for OTM operation. This w i l l s i g n i f i c a n t l y
reduce the total synthesis gas flow to AGR and the subsequent gas feed to the
Stretford sulfur recovery unit. In the Stretford unit, the bulk of the H2S is
absorbed, oxidized to elemental sulfur, and f i l t e r e d as a wet sulfur cake.

The gasifier blowdown and plant runoff are treated in a wastewater treatment unit
consisting of chemical treating and ammonia stripping prior to biological
t~'eatment.

3-83
WATER

C,RINOING ,~
I
GASIF~ATION~
I ~'o--,~G
WATER
}

[
~ I a~ pREPARATION

LOCK/~OPPE
I CLARIFJERI

AIR AIR i
SEPARATION OXYGEN
PLANT NITROGEN WASTE
I WATER
O0 TREATMENT
C(~t~VENT '~' COzto
UREAPLANT
SULFUR I
IS' rI
RECOVERY
I SULFUR

I c°.H,..
CONVERTER
COS
HYDROLYSIS
H I I
SYNGASto
AMMONIAPLANT

WATER

FIGURE 6

TVA AMMONIA FROM COAL PROJECT


OTM PROJECT

TVA is proposing to install a LPMEOHdemonstration unit at the Office of


Agricultural and Chemical Development's (OACD) Advanced Coal Gasification
Technology (ACGT) unit to demonstrate OTM production using fuel gas from a coal
gasification plant (Once-Through Methanol Project). The existing coal
gasification unit will be modified to feed a new 35-ton-per-day LPMEOH
demonstration unit. The combined ACGT f a c i l i t y and methanol demonstration unit
will test the system with a variety of coals under appropriate u t i l i t y operating
conditions. Slipstream tests of one or more gas-phase methanol production
processes may also be included. The results from this program will provide
u t i l i t i e s that are planning the construction of IGCC plants with sufficient
engineering data to include OTM capability. The primary test parameter is
operaLion on CO-rich, coal-derived synthesis gas. The four tests for the LPMEOH
process w i l l include:

Process variable test for gas composition, pressure, temperature,


superficial gas velocity, space velocity, and turndown.
Baseline comparison test between the LPMEOHdemonstration unit at
TVA and the LaPorte PDU (5 t/d).
Catalyst addition/withdrawal test.

Extended operating test for verification of catalyst l i f e and


catalyst deactivation.
The OTM project schedule is shown in Figure 7. The project is scheduled to begin
in October 1986 and to end in February 1990.

DESIGN MODIFICATIONS

In the original ACGTdesign, the CO-shift reactor converted a substantial fraction


of the COS in the synthesis gas to H2S. There was a concern about the COS
hydrolysis unit's a b i l i t y to handle the increased loading during CO-shift bypass
operations. However, recent tests during the ACP showed that the COS hydrolysis
unit can adequately convert the COS to H2S reaching near equilibrium conditions.

The Selexol AGR system was designed to remove CO2 from a shifted fuel gas with an
inlet CO2 concentration of 35 percent. Without CO-shift, the Selexol i n l e t CO2
concentration is expected to be 15-20 percent and the outlet to be about 5-6
percent. Increased CO2 slippage can be obtained in the existing Selexol AGR
system by both eliminating the solvent flow and increasing the solvent operating
temperature in the lower half of the absorber. This w i l l additionally reduce the
total gas flow to the Stretford unit.

3-85
DESIGN 26 MONTHS
~D
0o
o~
PROCUREMENT I I
18 MONTHS
I-.:
(.,1
SITE PREP 0 I-----I
3 MONTHS
CONSTRUCTION I I 0
(.,0 14 MONTHS
I
m
QO START-UP
6 MONTHS
OPERATION I
8 MONTHS
FINAL REPORT
13 MONTHS
I I I
1986 1987 1988 1989 1990

FIGURE 7
ONCE THROUGH METHANOL PROJECT SCHEDULE

lull IllU VIlli ..... L..... v r . . . . . . .


Although zinc oxide guard beds can remove sulfur to less than 1 ppm at 400OF,
other [race contaminants may require higher guard bed operating temperatures. In
addition to the existing ACGT zinc oxide guard beds, zinc oxide and alumina guard
beds will be installed with the LPMEOHdemonstration unit. These new beds will
have a fired heater to provide an operating temperature range up to 700OF. This
will allow for an evaluation of guard bed material, operating temperature, and
trace contaminant level on catalyst deactivation.

3-87
5. SUMMARY

IGCC f a c i l i t i e s are a clean, efficient, and economic means of generating electric


power from coal. A recent economic assessment by Fluor Engineers, Inc. (10)
provides an evaluation of 500-600 MWplants based on Texaco gasification and
state-of-the-art General Electric gas turbine technology. Although these
facilities would be equivalent in capital investment to conventional coal-fired
steam plants, Lhe resulting cost of power would be lower due to higher efficiency.
Operating f l e x i b i l i t y in an IGCC plant may be more limited than in a conventional
coal Fired plant. One efficient means of providing this operating f l e x i b i l i t y in
an IGCC plant is to convert some of the energy from the gasifier into a storable
liquid fuel using OTM synthesis. Several methanol synthesis technologies could be
used for this purpose. The LPMEOHprocess is being developed with this particular
application in mind. For instance, approximately 25 percent of the available fuel
gas energy could be converted to fuel grade methanol in a LPMEOHunit with a
capital investment of about 5 percent of the IGCC plant cost. This would allow
the gasification section of the plant to run at higher operating levels providing
baseline power generation. The stored methanol could then provide additional
energy for peak or varying load demands. ChemSystems is currently performing an
economic assessment of a commercial IGCC/OTMplant for comparison with a
standalone IGCC plant.

Liquid Phase Nethanol technology has been demonstrated at the five-ton-per-day


rate on synthetic CO-rich feed gas at the Air Products PDU in LaPorte, Texas.
While this program is continuing at Air Products, i t is now necessary to test with
coal-derived synthesis gas. The OTM project at TVA/Muscle Shoals w i l l provide
such a demonstration at a scale of operation needed for designing future
commercial-scale f a c i l i t i e s .

3-89
6. REFERENCES

I. Coal Gasification Systems: A Guide to the Status, Applications and Economics;.


Palo Alto, CA: Electric Power Research Institute, AP-3109.

2. E. Supp, "Technology of Lurgi's Low Pressure Methanol Process," Chemtech, July


1973, pp. 430-435.

3. Private communication with Lotepro Corporation, July 1985.

4. Private communication with Lurgi Kohle and Mineraloltechnik GmbH, August 1985.

5. W. Wolfer, "Helpful Hints for Physical Solvent Absorption," Hydrocarbon


Processing, November 1982, pp. 193-197.

6. "Katalco 32-4 Desulfurization Catalyst," Katalco Corporation brochure, Oak


Brook, I l l i n o i s 60521.

7. P. E. Jensen and K. Sondergard, " Carbonyl Sulfide Removal with Zinc Oxide
Masses," paper presented at AIChE meeting, Denver, Colorado, August 1983.

8. R, E. Smith, G. C. Humphreys and G. W. G r i f f i t h s , "Optimizing Large Methanol


Plants," Hydrocarbon..Processinq, May 1984, pp. 95-100.

9. E. Supp, "Improved Methanol Process,".Hydrocarbon Processinq, March 1981, pp.


71-75.

10. Cost and Performance for Commerical Applications of Texaco-Based Gasification-


Combined-Cycle Plants. Palo Alto, CA: Electric Power Research Institute,
AP-3486, April 1984.

3-91
LOW TEMPERATUREMETHANOLPROCESS: THE NEXT STEP

R. Sapienza
Brookhaven National Laboratory
ABSTRACT

'Although properties of methanol in a number of instances significantly d i f f e r


from those of conventional fuels and thus require some departures from standard
equipment and procedures, there appear to be no fundamental problems that might
impede the use of this extremely clean fuel in u t i l i t y boilers and gas turbines
were i t sufficiently available in quantities for widespread u t i l i t y use at a
competltive price". This quote from EPRI report AP-3342 demonstrates the promise
and the problem associated with methanol as a u t i l i t y fuel. The electric u t i l i t y
industry has generally perceived methanol as a synthetic fuel derived from coal,
oil shale and other mineral sources. But, the world's abundant natural gas
resources could provide methanol in fuel quantities to the u t i l i t y system.
Natural gas liquefaction is the current major option available for international
export transport of natural gas.
Gas production is on the increase and international trade even more so, with LNG
making most progress. It has been projected that by the year 2000, approximately
190 million (netric tons per year of LNG could be moving in worldwide trade with
delivery mainly to the heavily industrialized areas of North America, Europe and
Asia-Pacific.
The further penetration of natural gas into distant markets can be substantially
increased by a new methanol synthesis process under development at the Brookhaven
National Laboratory.
The new methanol process is made possible by the discovery of a catalyst that
drops synthesis temperatures from about 275% to about 100%. Furthermore, the
new catalyst is a liquid phase system, which permits the synthesis reaction to
proceed at f u l l y isothermal conditions, in contrast to the hot spots that prevent
currently available pelleted, solid catalysts from operating efficiently. There-
fore, the new low temperature liquid catalyst can convert synthesis gas completely
to methanol in a single pass through the methanol synthesis reactor. This charac-
t e r i s t i c leads to a further major improvement in the methanol plant. Atmospheric
nitrogen can be tolerated in the synthesis gas, and s t i l l the volume of gas fed to
the reactor can be smaller than the volume of gas that must be fed to the reactor
when accommodating the very low conversions furnished by the best of currently
available catalysts.
The energy disadvantage of the methanol option must be balanced against the advan-
tage of a much lower capital investment requirement made possible by the new BNL
synthesis. Preliminary estimates show that methanol conversion and shipping
require an investment for liquefaction to methanol, and shipping liquefied
methanol that can range from 35-50% of that needed for the LNG plant and LNG
shipping fleet. This large reduction in capital requirements is expected to make
liquefaction to methanol attractive in many cases where the LNG capital needs are
prohibitive. Alternately, the economically viable minimum size can be s i g n i f i -
cantly smaller for the methanol route, which should serve to expand markets
distant from the production areas. The specific design of either LNG or methanol
liquefaction and shipping w i l l vary with gas cost, but a significant increase in
gas markets may be expected from the introduction of the new methanol synthesis
process.
3-95
1. INTRODUCTION

T. E. O'Hare, R. S. Sapienza, D. Mahajan, and G. T. Skaperdas

IN~portant sources of natural gas are situated far from markets, and a significant
commercial activity has grown up based on liquefying natural gas at the source.
The liquefied Natural gas (LNG) is transported in specially designed LNG tankers
to distribution ports equipped with LNG vaporization f a c i l i t i e s which then deliver
the vaporized gas into a pipeline system for distribution to users.

CONVENTIONAL LNG

Modern LNG f a c i l i t i e s can deliver almost 80%-90% of the energy extracted from the
g~s well to the distribution pipeline, but the system requires very large capital
investrT~ent. Because evaporation losses of cryogenic LNG must be controlled, the
tanker Lransport fleet requires specially constructed, expensive ships dedicated
to this service alone. This necessitates a large transport shipping capital
investment which must be provided by the LNG project i t s e l f . In addition, a
dedicated receiving f a c i l i t y is required to revaporize the LNG adding to the
investment requirements for the project. The receiving f a c i l i t y is generally
located distant from normal shipping ports to avoid serious hazards associated
with possible s p i l l s of liquefied methane.

CONVENTIONAL METHANOL

Natural gas carl, and is, converted to methanol in large commercial plants using
well established technology. Though chemical conversion to methanol introduces
somewhat larger fuel value loss than does simple liquefaction of methane, the
magnitude of these losses depends on the technology used for methanol synthesis.
Further~ore, these conversion losses are not accompanied by losses due to evapora-
tion during transportation or during revaporization so that a methanol synthesis-
transportation-delivery system need not necessarily be uneconomic in competition
with LNG. indeed, the potential capital savings of a new methanol option would
indicate that the delivery of natural gas values to distant markets, by means of
methanol, merits careful consideration.

METHANOL PROCESSCONSIDERATIONS

Conversion of ndtural gas to methanol requires two process operations. Methane


must f i r s t be converted to synthesis gas, a mixture of carbon monoxide and
hydrogen, which is then converted to methanol in a second synthesis step.

3-97
The preponderant technology for making synthesis gas from methane now uses steam
reforming of natural gas. The steam reforming reaction is as follows:

CH4 + H20 ~ C O + 3H2 (1-1)


and has an endothermic heat of reaction of 88,500 Btu per mol methane at 64°F.

Additional methane or other fuel must be burned in the firebox of the reformer
furnace to supply t h i s steam reforming reaction heat and, in addition, the heat
needed to generate the steam required for the reforming
reaction.

Furthermore, steam reforming makes 50% more hydrogen than is needed for methanol
synthesis:

CO + 2H2 -----,-CH3OH. (1-2)


If there is a local market for extra hydrogen i t can be so used, but, i f not, i t
must be vented, resulting in an effective loss of methane, or i t may also be
converted to methanol i f inexpensive carbon dioxide is available at the site.
This reaction is:

1/3 CO2 + H2 -----~1/3 CH30H + 1/3 H20. (1-3)


An a l t e r n a t i v e solution to the excess hydrogen production of the usual steam h
reforming process is to combine steam reforming with a c a t a l y t i c autothermal reac-
t i o n , as proposed by Lurgi. This procedure does, however, require an oxygen
plant, adding to the complexity and investment for syngas preparation.

]n contrast, partial oxidation of natural gas yields an ideal methanol feed gas.
The reactions are:

CH4 + 1/2 02 ------~CO + 2H2 (1-4)

CO + 2H2 -----~-CH30H. (1-5)


These reactions are both exothermic so no fuel is required. Furthermore, all the
hydrogen is consumed in making methanol. In spite of these advantages of partial
combustion, steam reforming is being used for methanol synthesis gas and partial
oxidation is not. This is due to the need for oxygen in partial combustion when
the synthesis gas is to be used for conventional methanol synthesis. Oxygen is
required because methanol synthesis using conventional catalysts operates at a low
gas conversion and a large recycle stream must be used. Inerts that build up in
this recycle stream must be kept low for process efficiency, and this necessitates
the use of oxygen rather than air in partial oxidation to avoid excessive purging
necessitated by introduction of atmospheric nitrogen. The investment and opera-
ting costs for the oxygen plant, and for the power generation needed to operate
the oxygen plant, are so high that they overwhelm the cost of the extra natural
gas needed for reforming, as well as the cost of the expensive reforming furnace.

BROOKHAVEN METHANOL CATALYST

The methanol catalyst being developed at the Brookhaven National Laboratory has
been shown to provide very high conversions of synthesis gas. Conversions of 90%
and higher may reasonably be expected because the catalyst is very active, at ve~

3-98

]
low temperature, and i t is a solution so that reaction heat removal can be
decoupled from k i n e t i c s . Thus, optimum performance, chemically and thermally, may
be b u i l t into the system separately, each designed to optimize heat removal and
kinetics with l i q u i d catalyst c i r c u l a t i n g between the two sections. The a b i l i t y
to c i r c u l a t e catalyst also provides the opportunity to regenerate catalyst affec-
ted by temporary operating errors, while the plant continued to operate, obviating
the need to shut down for replacement of c a t a l y s t .

Thus, the new ~NL catalyst makes i t possible to take advantage of the improved
naLural gas requirements of p a r t i a l oxidation because the high conversion elimin-
ates the recycle stream, and thereby permits use of a i r rather than oxygen, saving
these large costs for oxygen generation. Furthermore, atmospheric nitrogen leaves
the reJctor at reaction pressure and can be expanded to provide energy for a i r
compression. As a r e s u l t , the new BNL catalyst makes i t possible to use a
synthesis gas preparation system in which large savings in investment and opera-
ting costs can be obtained. Thus, the prospects of using methanol, rather than
LNG, to transport natural gas values to market can be s i g n i f i c a n t l y improved in
comparison with steam reforming and conventional synthesis catalysts as now
practiced.

BROOKHAVEII METHANOL PROCESS

To explore this p o s s i b i l i t y , the flow sheet of Figure 1 has been prepared to show
the f a c i l i t i e s required to convert natural gas to methanol. The t o t a l operation
consists of four sections. F i r s t , natural gas is converted to synthesis gas using
c a t a l y t i c secondary reforming, with a i r y i e l d i n g a synthesis gas containing some
40% nitrogen, with the remainder being largely hydrogen and carbon monoxide in
about the needed 2 to I r a t i o . This operation is very s i m i l a r , though not f u l l y
i d e n t i c a l , to a well established technology used commercially in tonnage ammonia
plants for several decades. Second, the syngas is sparged through a reactor
containing the BNL catalyst solution, which is also circulated through a cooler to
remove the large heat of reaction. Third, the reactor vapors are cooled in stages
to recover a l i q u i d stream containing product methanol, the residual flash gas
becoming available as a fuel gas. F i n a l l y , a gas turbine using the flash gas as
fuel serves to compress the needed a i r , to recover the pressure energy in the
nitrogen of the flash gas and to convert part of the extra energy into e l e c t r i c
power.

Specifi,:aIly, the flow sheet of Figure 1 has been designed for reacting natural
gas with cur~ipressed, preheated a i r in the Syngas Reactor to produce synthesis gas
by the established secondary reforming reaction. The raw synthesis gas exits the
reactor at about 1660°F and is cooled in four steps, t r a n s f e r r i n g i t s heat energy
to feed a i r , flash gas, steam generation and f i n a l l y to cooling water. The
syngas, now at lOg°F, is dried in dessicant beds and fed to the synthesis reactor.
The c a t d l y s t , which is in the l i q u i d state, is held in a reactor, and dried
synthesis gas flows up through the reactor in intimate contact with the catalyst
The synthesis reaction:

CO + 2H2 ----"CH30H (1-6)


produces methanol and a substantial amount of heat. To maintain the desired reac-
tion temperature of 248°F (120°C), a stream of catalyst is circulated between the
reactor and an exchanger, which rejects heat of reaction to cooling water. At the
temperature and pressure chosen for reactor operation, namely 248°F and 194 pounds

3-99
per square inch absolute, methanol produced s y n t h e t i c a l l y is carried overhead in
the spent gas, but methanol present in the catalyst bath remains in the reactor.
This d i s t r i b u t i o n of methanol is controlled by adjusting the temperature and
pressure in the c a t a l y t i c reactor.

The spent gas is then cooled in steps to 25°F, f i r s t by cooling water, then by
flash gas, and f i n a l l y by r e f r i g e r a t i o n , and enters a flash drum. Raw methanol is
recovered from the two flash drums as l i q u i d , and is d i s t i l l e d in two towers to
remove l i g h t components in the methanol s t r i p p e r and heavy materials in the
methanol d i s t i l l a t i o n tower, to produce product methanol for storage and sale at
the rate of 3,000 metric tons per stream day.

Flash gas separated from raw methanol in the flash drum absorbs some heat, f i r s t
from the flash gas, and then the synthesis gas to be heated to 600°F. This
temperature destroys undesired components present in the flash gas, and the
products of t h i s operation are trapped out of the flash gas in an adsorber. The
resulting cleaned flash gas is burned in the combustor of the gas turbine and the
combustion gas is expanded to generate power for the a i r compressor. Significant
excess of power is also generated.

Thus, the integrated plant converts natural gas to methanol and requires only a
catalyst, b o i l e r feed water and a cooling water supply. All the energy required
to operate the plant and deliver product methanol is obtained from natural gas.

The plant of Figure i can be operated in a variety of manners. For instance, the
size of the synthesis reactor, the flow rate of syngas, and the reactor tempera-
ture and pressure, determine the extent of conversion of synthesis gas. The
temperature chosen for operation of the second flash drum determines the degree of
recovery of methanol from the flash gas. As t h i s temperature is raised, more
methanol escapes into the flash drum to raise the heating value of the gas turbine
fuel gas, and thus to generate more power. The effect of these variables is
i l l u s t r a t e d in Table I , in which conversions of 70% and 90%, and flash tempera-
tures of IO0°F and 25°F are indicated for a unit output of 3000 metric tons per
stream day.

Table I

BNL METHANOL PROCESS EFFECT


OF SELECTED DESIGN VARIABLES

Methanol Produced MT/SD 3,000 3,000 3,000


CO Conversion % 90 90 70
Flash Temperature °F 100 25 100
Natural Gas Feed mol/hr 9,735 9,220 12,480
Flash Gas HHV Btu/SCF 42 27 90
Combustion Temperature °F 2,180 1,778 3,150
Net Power Output BHP 22,900 13,000 238,000
Catalyst Volume Gallons 110,000 II0,000 52,000
Flash Refrigeration Power BHP 0 720 0
Fuel Value Recovery % 67.9 71.8 53.3
Energy Recovery % 69.4 72.6 65.5

3-100
At 90% conversion and 25°F flash temperature, the methanol product has 71.8% of
the heatiny valu~ of the natural gas feed. In a d d i t i o n , 13,000 brake horsepower
of e l e c t r i c energy (9,700 KW) are available from the gas t u r b i n e , a f t e r allowing
for a i r compression and r e f r i g e r a t i o n power. I f t h i s energy, expressed as heat,
is included, the energy recovery becomes 72.6%.

If r e f r i g e r a t i o n were not used, the methanol loss is more important than the addi-
tional power generated by burning methanol in the t u r b i n e combustor, and the fuel
energy recovery and fuel and power energy recovery drop to 67.9% and 69.4% respec-
tively. On the ,]Lher hand, the heating value of the flash gas fed to the combus-
tot rises fro,~ 27 to 42 Btu per standard cubic f o o t . I f the combustor should,
however, require the higher heating value, i t would be b e t t e r to use supplemental
natural gas and save the methanol.

If the c a t a l y s t volume were reduced from 110,000 to 52,000 gallons, the conversion
of carbon Hono~ide would drop from 90% to 70%. The extra syngas l e f t in the flash
gas would raise the flasll gas heating value to 90 Btu per standard cubic f o o t , but
the fuel energy and t o t a l energy recoveries would drop to 53.3% and 65.5%, respec-
tively.

It seems clear that an adequate catalyst volume, and a thorough recovery of


product nethanol from flash gas are clearly warranted by the additional capital
required for these f a c t o r s .

CAPITAL <EL~IJIREMENTS - BROOKHAVENMETHANOL OPTION

In ; r d e r to ew~luate possible economic advantages of the BNL methanol synthesis


the equipment shuwn on the flow sheet of Figure 1 was used as a basis. All the
pieces of ~rocess equipment shown on the flow sheet were sized f o r operation to
provide 90; conversion of carbon monoxide and recovery of methanol from flash gas
cooled to 25°F. In addition to the process sizes and d e s c r i p t i o n f o r each item of
equil~m~,~t, the materials of construction needed f o r adequate equipment l i f e were
specified. So~.e flow sheet functions were performed in a single u n i t and some in
m J l t i p l e u n i t s . All pumps were spared and, in a l l , 90 pieces of equipment were
s~zed and t h e i r costs estimated. From these pieces of process equipment bulk
~te,~s such as piping, s t r u c t u r e s , b u i l d i n g s , concrete foundations, e l e c t r i c a l gear
~rl:I instruF,ents were estimated f o r the p a r t i c u l a r array of process equipment
required. To t h i s subtotal were added f i e l d erection costs at the plant s i t e ,
boche of fie,' engin,=ering COSTS, and appropriate o f f s i t e s to a r r i v e at a t o t a l
i n s t a l l e d cost f u r an erected plant ready to operate. All costs were based on a
,}.S. Gulf C,)ast ],]cation.

The esti(~ate thus obtained shou]d, therefore, serve to provide a dependable


projection of ,~ethanol costs using the BNL process assuming that further pilot
plant develop~nent indicate neither improvements nor degradation in performance now
~,rojected from the research work. In addition, the distribution of costs among
~he various portions of the plant that may be derived from the estimate can serve
to olJtimize prucess conditions for minimum plant cost in subsequent designs. I t
may be noted that such optimization, beyond that shown in Table 1, was not
included in the flow sheet of Figure 1, which was designed before the estimates
were a v a i l a b l e .

3-101
COMPARATIVE PROJECTED COSTS

The results of the cost estimating studies are shown in Table 2, which sets forth
investment and operating costs for three options. First the BNL methanol process
investment derived from the estimating work described above is presented together
with capital requirements for shipping and receiving terminal f a c i l i t i e s . The
basis of the comparison is a methanol production of 15,000 metric tons per day of
methanol in a plant consisting of 5 trains, each producing 3,000 metric tons per
day. In addition, a plant estimate for similar f a c i l i t i e s based on a modern,
conventional, low pressure process is included. These costs were estimated in the
same manner by the same team as the costs for the BNL methanol process, and may be
expected therefore, to represent a dependable comparison. The conventional
methanol plant, not enjoying the merits of decoupling kinetics and cooling in the
synthesis steps, required 7 trains to reach 15,000 metric tons of methanol per
day.

Finally, costs for LNG f a c i l i t i e s to define the same fuel energy as 15,000 metric
tons per day of methanol were determined from published information for LNG
facilities.

All the investment and operating cost information in Table 2 were prepared by the
Stone and Webster Engineering Corporation.

The strong cost advantage provided by the BNL methanol catalyst and process is
shown by the capital requirement for the production plant, and offsites needed to
support the production plant. The BNL methanol process requires an investment of
$518 million, compared to $1066 million for a modern, conventional, low pressure
process. Clearly, the large potential capital savings j u s t i f y continued develop-
ment of the process. The total investment for shipping and receiving 15,000 tons
per day of methanol are $904 million and $1452 million for BNL and conventional
methanol respectively. Operating costs result in a delivered fuel cost of $3.59
per million Btu for BNL methanol as against $4.59 per million Btu for conventional
methanol. While most of the BNL savings derives from the capital charges, a
significant saving is also provided by the lower natural gas requirement that
results from the more efficient syngas manufacturer by means of secondary cataly-
t i c reforming with air.

Clearly, the BNL methanol process offers a large cost improvement over current
technology. The advantage is of such magnitude that previous comparisons of
methanol versus LNG for shipping gas values need no longer prevail.

The relative positions of BNL methanol and LNG are also illustrated in Table 2.
The capital requirement for BNL production and offsites is $518 million compared
to $724 million for LNG. The advantage of BNL methanol is again large. But when
shipping and revaporization are included the advantage is much larger, BNL requir-
ing $904 million vs. $1,720 million for LNG, the incremental LNG investment of 40%
in the process plant becoming 90% for the entire f a c i l i t y . Finally, the delivered
fuel cost is estimated to be $4.80 per million Btu for LNG compared to $3.59/
~,iIlion Btu for BNL.

3-102
Table 2

DELIVERY OF FUEL VALUES

Delivered Fuel Methanol Methanol LNG


Process Brookhaven Conventional Mixed
Low Pressure Refrigerant
Fuel Delivered, Metric Tons/Day 15,000 15,000 7,375
Billion Btu/Day, HHV 323 323 323
Number of Production Trains 5 7 2
Number of Shipping Tankers 4 4 4
Shipping Distance, Miles 6,800 6,800 6,800
Capital Costs, M i l l i o n $

Production Facilities & Offsites 518 1,066 724


Shipping Tankers 316 316 721
Receiving Terminal 70 70 275
Total Capital 904 1,452 1,720
Operating Costs,.. M i l l i o n $ Per Day

Production 183.1 196.8 166.6


Shipping 72.5 72.5 80.2
Receiving 0.8 0.8 6.8
Capital Charges, 15% Per Year 135.6 217.9 258.0
E l e c t r i c i t y Credit @ 1~/kwh (9.3) . . . .
Total Costs 382.7 488.0 511.6
Delivered Fuel Cost S/Million Btu HHV 3.59 4.59 4.80

It may be noted, however, that the fuel capacity of Table 2 requires 5


parallel trains for BNL methanol compared to 2 trains for LNG. In fact the
capacity of Table 2 is at the low end of the range for LNG and i t is useful to
compare BNL methanol with LNG at the upper end of LNG plant capacities. Table 3
compares LNG bdsed on a capacity of 2 b i l l i o n standard cubic feet per day of LNG
(equiwIent to six times the capacity of Table 2) with methanol at 15,000 metric
tons per strea,~l day, namely the capacity of Table 2. In addition, the shipping
distance in Table 3 is reduced to 3,000 miles to minimize shipping costs and
valJorIzdtion losses for LNG.

As shown in Table 3, a large LNG plant enjoying the advantages of scale and the
shorter shipping distance of 3,000 miles has a delivered fuel cost at the terminal
of $2.98/milllon Btu of higher heating value, which is the same, e s s e n t i a l l y , as
the delivered fuel cost of the much smaller BNL methanol f a c i l i t y . The BNL
process, one may conclude, makes available at competitive costs much smaller gas
reserves, requiring very much smaller capital investment for shipment competi-
t i v e l y to distant markets, and c l e a r l y appears to be the choice over LNG as soon
as further development of the process confirms present laboratory data.

3-103

I
Table 3

DELIVERY OF FUEL VALUES

BNL Methanol LNG


Delivered Fuel, B i l l i o n Btu/D 323 2,030
Number of Production Trains 6
Number of Shipping Tankers 27
Capital Investment, $ M i l l i o n
Production & L}ffsites 518 2,675
Shipping Tankers 158 2433.5
Receiving Terminal 7O 600
Total Capital, $ Million 7-4-d 5 ~
Operating Costs, $ M i l l i o n Per Year
Production 183.1 827.9
S~i ppi ng 36.2 270.7
Receiving 0.8 42.8
Capital Charge 111.9 856.0
E l e c t r i c i t y Revenue (9.3)
Total 323 1997.4
Delivered Fuel Cost S/Million Btu HIIV 3.02 2.98

F'IA
Flash Drum
E-6
Flash Refng.
Cooler F'I
Flash Drum

~ R-I
Syngas
RlO~lof ;~
I I ~"
T.2
o,,tl I ,J~

E-I
P
~'l ] I I f---~ P-JAaS
R-2
Syrll~eSll
R~ctor
6.~S Tacbtr~ Rice|or /fu
C,~mNIt~ C~ollr
I

®
Synga,
Dryer,

I
.i. !i!,:.
&-

3.000 M.TON/DAY METHANOL NATURAL GAS FEED

Figure I .

3-104
RESEARCH ON ONCE-THROUGH METHANOL

S. Lee, V. Parameswaran, A. Sawant, M. Ko and D. H. Cho

University of Akron

:2

d
A I; S [' RA C I ~

Itn~Lnee~ tng and fundamental data for the catalytic methanol


synthesis have recently been obtained to support the process
development and design of the Liquid Phase Methanol Synthesis
Process (LP~IeOH). Even though various factors and concerns directly
,:,r indirectly related to the process economics have been
iI,vesti£ated, there still exist many questions to be answered, viz.,
r,,]e of carbon dioxide and water in the synthesis, various causes of
c,~talyst deactivation, mechanistic and microscopic behavior of the
e;]ta[yst, transport mechanisms, etc. Most of these subjects are not
,:,,Iv apI,licable to the Liquid Phase Methanol Synthesis Process, but
al~,~ t,, the catalytic synthesis of methanol, in general, from
syn~as, Without clear understanding of such scientific and
en£ineer i,,g problems, truly innovative breakthrough in the process
c,)i,uept would be difficult to attain.

As a p,~rt of serial efforts made, this paper focuses on the recent


d~vclc, p~u~,nts and findings in the catalytic, once-through synthesis
of nlcthanul from syngas, and consists of four major parts: (a) brief
review of prior achievements, (b) role of carbon dioxide in methanol
sv,,thesis, (c) use of BASF S.3.85 catalyst, (d) microscopic or
mechanistic phenomena in the methanol synthesis catalyst.

1 NTIR()I)Ut:T I ON

In tl,e catalytic synthesis of methanol from syngas, one of the few


lavt:s l:l,aL are generally agreed upon without controversies is that
the, over~Jl] synthesis reaction is exothermic. On the reactor scale,
the exothermic heat of reaction can cause a thermal instability
pr,,h]em whenever the 'slope' condition for stability is violated.
Even on the catalyst scale, the exothermic heat of reaction can
,:au~u ~ru~lems of sintering, fusing, annealing, and/or local hot
~p~,t~, ,I] of which are undesirable from the standpoint of catalyst
]ilc.

l~n ti~e ,~nvel Liquid Phase blethanol Synthesis Process, the catalytic
synthesis of methanol from syngas is carried out in a three-phase
mo,lc, tl~us alleviating the problems resulting from the exothermic
hu,t nf reaction. The improvement made in the thermal behavior of

3-107
t h t ~; ct,emi, c a l system is due to the existence of a thermally stable
;JFi ~I cl,emically inert oil which has an inherently higher thermal
ulJC r l . : i ,~ than the gas phase in conventional vapor phase processes.
[]OWr: V¢' r , this introduction of the liquid phase into the catalytic
,:hemic,~l system spawns other new problems that are either more
:3~VC'F~ or have not been observed with the vapor phase processes.
i-iu c h p~:ob]ems include:

;~,dditional mass transfer limitations and


( b ) greater difficulty in u n d e r s t a n d i n g the c h e m i c a l and
Izransport mechanisms. As with other catalytic
l,rocesses, the life of the c a t a l y s t in the n e w ± y
developed ~rocess is also of m a j o r c o n c e r n as it
d~rect~y affects the economic feasibility of the
flrocess.

~W C'V t~C the following technical assessments can be made about the
q ,i d P h ase Methanol Synthesis Process:

Due to the thermally stable nature, the process


c a n b e operated at h i g h e r t e m p e r a t u r e s t h a n the v a p o r
phase process, without causing reactor instability
and/o r catalyst annealing.
W i t h o ut considering, chemical equilibrium, higher
tempe rature operations would normally represent
highe r reaction rate.
The exis tence of an inert oil phase creates
additional mass transfer resistances, such as
gas-] Jquid and liquid-solid mass transfer. Such
ceslstance s can limit the o v e r a l l p r o d u c t i v i t y per
r e a c t o r vo lume.
% he in e r t o il chosen m u s t p o s s e s s the c a p a c i t y to
dissolve rea c t a n t gases (hydrogen, carbon monoxide
and carbon d ioxide) selectively, have a high boiling
point and e chemically inert. F r o m the e c o n o m i c
~ t a n d p o i nt the oil should be inexpensive and
,;eusabJe. '
IIany causes for catalyst deactivation h a v e been
identified and r e s o l v e d . T h e r e is s t i l l good r e a s o n
to believe that more research to increase the
c a t a l y s t life is e s s e n t i a l .

T ,~ ! it,al d e c i s i o n to implement the liquid phase process instead of


:~,,v of the other vapor phase processes would depend primarily on:
(,) ir,proved productivity per reactor (volume); (b) enhanced
,: ~ a ! y ~ r life or applicability of superior catalysts; (c) increased
p~ r-p~,~s conversion to methanol; and (d) discovery of more
;i,e×p,:r, sive but better oils. To achieve these goals, first of all,
I~l~c l,r,-,cess chemistry as well as transport mechanisms must be
clear]y ~Jnderstood.

3-108
I'1, u f ,-,
,,wing r e s e a r c h work has been performed at The U n i v e r s i t y of
,\:(I ,11|~
~, order to support the development of the Liquid Phase
~l, I tJan,,
S y n t h e s i s Process and to improve the general understanding
Of c d k 2* k:ic m e t h a n o I synthesis:

1)
pceliminar~ range finding calculations for heat and
~ass trans er r e s i s t a n c e s
(2)
thermodynamics: phase and chemical equilibria
(3) thermal stability analysis
(4)
:iJ,Lrinsic kinetics and development of a lobal
,:eaction rate e x p r e s s i o n g
5)
~l:i ~fusional kinetics and the analysis of pore
,l:iffusional [imitation
6)
r~e;,suremenLs
un,I anal),sis of external mass transfer coefficients
7
cle aarificat:ic;n
d ctivation of potential causes for catalyst

testing other catalysts for the same process concept


u,l,lerstandJn
,~:icroscop[c }i e~o fe l . c h e m i c a l and transport mechanisms at

I~',:',~ u ] k s
I cc,m tasks I to 5 have been summarized and recommendations
lJ,Lvc, bec
,, made in r e f e r e n c e [I]. Novel ideas and preliminary data
1! u r t he
analysis of mass transfer resistances in a slurry reactor
SV:. I. CI1] (
task 6), h a v e been summarized in r e f e r e n c e [2]. A complete
t] II~l ] \ ' s i s
of mass transfer with experimental data will soon be
54 Illilll}~t F J ,;f ~'t]
in a parallel publication. A study on potential catalyst
duv, I c l a t i,,n
has also been published in r e f e r e n c e [2] and more work
i '÷: C ~1 i f r c i i I I y
underway to provide additional confirmatory evidences.
k he current paper is d i r e c t e d towards the following:

I I) i~, u n d e r s t a n d the role of carbon dioxide in the


~ y n L h e s i s o[ m e t h a n o l from syngas.
(2) f,, present the
e x p e r i m e n t a l data with a BASF catalyst
~,, c o m p a r i s o n to EPJ catalysts.
t ~) i,, su~lmu r i z e the experimental findings on the
'-'~La 1 - V s t behavior, deformations, chemical and
"~'rl"hblogical changes, etc.

tt,,.~e newfindings, the current paper will s u m m a r i z e


FCqC',t l,Zh
the
findingsso f a r made in order to improve
the o n c e - t h r o u g h
rIIC I h,l11L, l
: vnl:hesis prucess in the liquid phase.

3-109
i
ROLE OF C A R B O N DIOXIDE IN M E T H A N O L SYNTHESIS

Background

]n the catalytic synthesis of methanol, there are only a few


cl,emica] species mainly involved. However, many researchers (Klier
and co-workers [3]; Kagan and co-workers [4,5,6]; Rozovskii [7];
Kuznetsov [8]; Kung [9]) have suggested conflicting explanations
regardinR the role of c a r b o n dioxide in the catalytic synthesis of
methanol from syngas. An extensive review on this controversial
~,,bject h;3s been made by W e n d e r and Sayari [10].

A key :issue in this c o n t r o v e r s y has been w h e t h e r methanol synthesis


proceeds through CO or CO 2. Depending on the c l a i m or p o s t u l a t e
that each researcher has proposed, even more controversies and
d:isagre,~ments have been generated on r e l a t e d issues, such as the
secondary reaction, adsorption mechanisms, role of ZnO, active form
of copper, etc. It s h o u l d be c l e a r l y noted that all these studies
have sn l~ar been made in v a p o r phase reactor operations. However, it
ix also believed that the c h e m i s t r y of m e t h a n o l synthesis is still
the same in the liquid phase mode, since the inert oil used in the
]~quid phase synthesis does not participate in any of the chemical
reactions (Lee [1]).

ljiLdc'FS t ~F~diFLg the true chemistry, however, becomes even more


s,~nJ Licant in liquid phase methanol synthesis. This is because
there, exist more operating parameters to be o p t i m i z e d in the liquid
phase process due to the e x i s t e n c e of the inert oil phase. Crucial
quenLior, s that must be answered for the liquid phase synthesis
proces~ are:

(a) At what level of c a r b o n dioxide concentration would


the m e t h a n o l p r o d u c t i v i t y be m a x i m i z e d ?
(b) Wl,at is the role of. w a t e r in the c a t a l y t i c s y n t h e s i s
oI~ m e t h a n o l ? Does w a t e r i n t e r a c t c h e m i c a l l y with the
catalyst inRredients, especially with catalyst
support A 1 2 0 3 ?
(c ) How is the produced water transported from the
c~talyst active sites to the bulk v a p o r phase of a
rr,actor? Can w a t e r b u b b l e s grow in the m i c r o p o r e s of
the c a t a l y s t ?
(d) Is it p o s s i b l e for c a t a l y s t c r y s t a l l i t e size to grew
h y d r o t h e r m a l l y in the p r e s e n c e of w a t e r ?

3-110
(e) Are any of the principal or intermediate chemical
species involved in catalyst crystallite
stabilization or deactivation? Is there any clue as
to the
size? cause of the growth in catalyst crystallite

Thezefore, the objective of this part of the investigation has been


to clarify the role of carbon dioxide in the Liquid Phase Methanol
Synthesis, with the aim of improving the process characteristics as
well as better understanding of the chemistry. This investigation is
al~o meaf~t to clarify some of the controversial issues presented in
the c u r r e u t literature.

Apl, roaches

g,:,t;~ micr,,scopic and macroscopic approaches are adopted. Macroscopic


~nvestig,~tion requires a series of reactor operations, whereas
mi~,,~scop:ic investigation ~ocuses on the molecular, catalytic,
c, vstalljne, adsorptive, or desorptive phenomena.

A ,,,,e-liter, mechanically agitated, autoclave-type, slurry reactor


h,,x been used for macroscopic investigation. The experiments have
he(,, made following the procedure of elimination, i.e., in a manner
of minin1~zing confounding effects. Detailed descriptions of the
slurry reactor system and the experimental procedure have been
published elsewhere (Lee [i]; Lee, et al., [2]).

The key chemical reactions involved in the controversy are as


f! o ] ] o w s :

klf
C(-). + 3 H2 ~=~ CH30 H + H20
kl b .(1)
k2f
CO~ + H2 ~=~ CO + H20
k2 b (2)

k3f
CO + 2 H2 ~=_~ CH30 H
k3b (3)
k4f
2 CO ~=-~ C + CO 2
k4 b (4)

3-111
From the kinetic standpoint, it does not appear to be wrong (at
least macroscopica]ly) to assume that all four chemical reactions
t~:~ke place at certain rates. It is o b v i o u s that only three reactions
(one of which must be (4)) out of the four are s t o i c h i o m e t r i c a l l y
independent. Since the carbon in (4) is not involved in the other
reactions, reaction (4) (carbon deposition or B o u d o u a r d reaction) is
taken separately. Anyway, such m a c r o s c o p i c investigation is s o m e w h a t
weak in making conclusions in terms of p o s s i b l e intermediates and
atomic exchanges taking place at c a t a l y s t active sites.

CO-Free E×periments

A series of kinetic experiments were made with the slurry reactor


using the EPJ-25 catalyst with a feed gas a b s o l u t e l y free of CO. The
feed gas composition was H 2 63.3 %, CO 2 31.7 %, CH 4 5.0 %. The
following hypotheses have been tested with respect to the
experimental data obtained under various temperature and pressure
conditions.

Hypothesis (I-l). The forward reaction of (I) is faster than the


forward reaction of (2) and m e t h a n o l is p r e d o m i n a n t l y produced via
direct hydrogenation of CO2, when CO 2 + H 2 is the feed gas
mixture. Such case is when klf > k2f with no special restriction
on kBf , except that it is of the same order of m a g n i t u d e as klf.
(Even this mild restriction becomes unnecessary when klf >> k2f. )

Hypothesis (I-2). Methanol is p r e d o m i n a n t l y produced via a 2-stage


mechanism of (2) and (3), when CO 2 + H 2 is the feed gas mixture.
Such is the case when k2f >> klf and k3f > klf.

Hypothesis (I-3). Reactions (i) and (2) are c o m p e t i n g when CO 2 +

11o is the feed gas mixture. Reaction (2) further goes to Reaction
(3) producing CH3OH.

Each hypothesis can be separately characterized in terms of


anticipated results on the m a c r o s c o p i c level. The results expected
when each hypothesis is a s s u m e d to be true are listed below and will
be compared with direct experimental findings.

3-112
Anl ici?at~,l l~esults from H y p o t h e s i s (I-l):
I) Verv liLtle CO will be d e t e c t e d in the exit stream of
the- reactor

J) ~he ,,umber of moles of water collected will be


appruximatelv 1 to the number of moles of
meth,~[to] p ~ o d u c e ~ qua
3) ]~he t-ate of disappearance of hydrogen will be
approximately equal to three times the rate of
methd.ol pro~uctzon
4) ] ' h e r~Jte of methanol production is a p p r o x i m a t e l y
equal, to the rate of c a r b o n d i o x i d e reaction
I
5) ]'he r~te of w a t e r p r o d u c t i o n is equal to the rate of
carb~,n dioxide reaction. However, this is not a
characterizing c r i t e r i o n Zor the h y p o t h e s i s and only
serve~ as an z n d i c a t o r of a good e x p e r i m e n t .

Af, t i c , p a t u , , I Results from H y p o t h e s i s (I-2):

(l Substantial a m o u n t of CO will be d e t e c t e d at the exit


strean~ of the r e a c t o r
(2 The ,,~Jd~ber of m o l e s of w a t e r p r o d u c e d is equal to the
sum ,,f the n u m b e r of m o l e s o~ m e t h a n o l p r o d u c e d and
the number of moles CO produced but not f u r t h e r
converted.
(3 l'he ~,un~ber of moles of water collected is greater
than the number of m o l e s of m e t h a n o l
(4 The l,vdrogen reaction rate is g r e a t e r than three
times the m e t h a n o l p r o d u c t i o n rate, i.e.,
rH2 3 rCH30 H .
(5) The ,ate of w a t e r p r o d u c t i o n is equal to the rate of
carb,,n d i o x i d e r e a c t i o n . H o w e v e r , this does not serve
as a c h a r a c t e r i z i n g , c r i t e r i o n for this h y p o t h e s i s ,
nJnr,, this holds for n y p o t h e s i s (I-l).

Anl: c i pat~!,,I k~ ~ s u l t s from H y p o t h e s i s (I-3):


(~) The results must be o b t a i n e d algebraically based on
comp~!'Lin~ r e a c t i o n s and e x t e n t s ol r e a c t i o n s .

In Tab es 1 and 2, hypotheses (I) and (2) have been tested with
,'xpe,:in,~ ntall/ obtained data. In order to make unbiased conclusions,
th~ tenbperatu~ has been varied between 210 and 250 C and the
p r e s s u r e between 35 and 70 atm.

As shown ~ ~, the tables, hypothesis (I-2) generates results


consistent witl, the e x p e r i m e n t a l observations. It is, h o w e v e r , still
difficult to c o n c l u d e that methanol is p r e d o m i n a n t l y produced via a
2-stage reaction scheme when a syngas of C02:H 2 = 1 : 2 is used
w~thout CO, since the m o l e c u l a r behavior of i n v o l v e d species on or

3-113
TABLE 1
HYPOTHESIS TESTING FOR (I-i)

~un T P YCO rH20 rH2 rMeOH rH20

lID C Atm YMeOH rMeOH rMeOH rc02 rc02

VP05 210 70 0.45 1.64 4.2 0.66 1.07


VP07 225 35 I.i0 2.27 5.1 0.48 1.08
VP06 237 35 1.54 2.62 5.1 0.42 1.09
VP08 237 50 1.06 2.07 4.7 0.46 0.95
VP04 250 70 0.90 1.92 4.4 0.51 0.98
V 71 i di t v NO NO NO NO YES "~

4~ Tl,Js does not serve as a criterion for the hypothesis;


but as an indicator of a good experiment.

TABLE 2
HYPOTHESIS TESTING FOR (I-2)

Run T P YC0 rH20 rH20 rH2 rc02

ID C Atm YMeOH rMe0H+rCO rMe0H rMe0H rc0+rMe0H

VP05 2]0 70 0.45 1.13 1.64 4.2 1.05


VP07 225 35 I.I0 1.08 2.27 5.1 1.00
VP06 237 35 1.54 1.03 2.62 5.1 0.95
VP08 237 50 1.06 1.00 2.07 4.71 1.05
VP04 250 70 0.90 1.01 1.92 4.41 1.03
V~]idity YES YES YES YES YES

uear the active sites are not considered in any of the above
hypoLheses. However, it still can be safely said: "if CO 2 + H 2 is
the feed gas mixture, hypothesis (I-2) is consistent with
macroscopic kinetic results". This conclusion should not be extended
to other feed gas conditions, since different compositions,
including the existence of CO, change the equilibrium nature of the
~ystem which, in turn, makes differences in mechanistic reaction

C02_Free Experiments

A series of kinetic experiments have been carried out with CO-rich,


gaseous feed composition was kept at
COo-free syngas. The
I0. It has been observed that the
H2:CO:CH 4 = 35 : 55 :
more precisely, the methanol productivity
catalytic activity, or
the reactor is operated in a continuous
drops consistently when

3-114
mode It should be mentioned that the reactor and catalyst were
protected against potential carbonyl contamination by the use of
activated carbon and molecular sieves The experimental results of a
contlnuous experiment zs shown in F i g u r e 1

The trend of decreasing catalytic actlvlty in terms of m e t h a n o l


production was observed for 72 hours on a c o n t i n u o u s mode with
samples belng
drawn every one to three hours After thls i n l t l a l
72-hour period
was over, the feed gas was s w i t c h e d to the o r i g i n a l
normal syngas The m e t h a n o l p r o d u c t i o n rate r a p i d l y i n c r e a s e d from
the 9 9 % level of the original rate to the 80 % level The
resumption shows that the decrease in rate during the C02-free
period was not totally due to deactivated catalyst This also
implies and supports that CO 2 does partzclpate in the vital
mechanistic steps of surface reaction zn m e t h a n o l synthesis Also
znterestlng is the fact that there seems to exist some unrecoverable
loss in the catalytic activity due to the 72 hour exposure to a
C02-free environment

Measurable amounts of CO 2 were still detected in the product gases


leavlng the reactor during the C02-free experiments Since the
feed syngas did not contain any measurable CO 2 durlng the
C02-free run, the source of CO 2 must have been from CO via a
chemical reaction In such case, the only conceivable chemical
reaction zs carbon deposition reactlon (Boudouard reaction),
2 CO = C + CO 2 It is also probable that the carbon deposited on
the catalytic sltes may have affected the catalytic activity The
permanently unrecoverable loss zn actlvlty even after the normal
syngas may have been due to carbon deposition on the catalyst
surface

There have been confllctlng opznlons expressed regarding the reduced


catalytic activity when the reaction ms c a r r i e d out in the absence
of CO 2 Klzer et al [3] believed that syngas mixture of H 2 + CO
overreduces the catalyst from Cu +I to Cu ° and the overreduced
catalyst does not possess the initial high activity In o r d e r to
check this possibility, the catalyst was slowly reoxldlzed by
passing air into the slurry reactor and then reduced again following
the standard ln-sztu reduction procedure (Lee, et al [I]) After
the reoxzdatlon and the subsequent reduction, the catalytic activity

3-115
0
o

EXTENDED EXPERIMENT TO
++.1+4. m
OETERMI NE THE INFLUENCE /
,(D.
OF CRRBON 0 10X 1 DE
/
I~ 4.

,~'~o~"~,:
~ ~."'~"= ~o° ~y~)o,
NORHRL FEED N]TH C02
o

FEED GR5 H JTttOU7 C02


rl" 4.

~oJ

FEEO CHANCEO
~. v
I I I
oo lb.oo 2b.oo 3b.oo ,=b.oo sb.oo eo.oo 70.00 8 0 . O0
TIHE ELRPSEO (HOURS)

Figure 1.
Q
¢3

¢:1,

n-ca
-r-¢~
EXPER I NENTS WI THOUT C02
¢_1
=o==
$
IN THE FEED GRS
(FEEB ORS W I T H H I G H CO:H2)
iT = 2 3 7 C. I~ = 9 3 0 1=SlO.)
Lo
I
++4-4. 4. 4.
4.
,~l

W= t 4- 4- 4. 4. ÷
4- 4.
4.
4.
4. 4. 4.
4. 4. 4.
¢~ 9 4.
41-
I--
b.I

== 4-

00 tO, DO 20,00 30.00 It0,00 50,00 60.00 "/0, O0 80


TIME ELRPSED {HOURSI • O0

Figure 2.
was rechecked and compared with the a c t i v i t y before this treatment.
It turned out that the a c t i v i t y before and after this treatment was
equal within experimental errors. Therefore, there is no strong
reason to believe that the catalyst subjected to a C02-free
environment deactivates due to o v e r r e d u c t i o n .

Based on this continuous run study (Figures i and 2), the following
conclusions can be drawn:

(i The presence of CO 9 in the syngas m i x t u r e means


mOrereactions~tha carabonSimpledioxldeParticipati°nis
a very importantmn
stoichiometriCspecies
that must be p r e s e n t for m e t h a n o l s y n t h e s i s i n v o l v i n g
the use of E P 3 - 2 5 type catalysts.
(2 ~ny_ syngas mixture that does not contain CO
ould narm the cop based EPJ c a t a l y s t The los~
activity is not zu~ er
ly recovered even after the
gaseous e n v i r o n m e n t is corrected. No clues were found
to support the possibility of o v e r r e d u c t i o n to be
blamed ~or the reduced activlty.
(3 Under COo-free, C O - r i c h s y n g a s conditions~ evidence
concerning the possibilSty of carbon qeposition
reaction was obtained, it the carbon deposition
reaction occurs under normal syngas c o n d i t i o n s , it
would affect the c a t a l y t i c activity.
(4 There seems to exist an optimal c o n c e n t r a t i o n of
COo, both for optimal CH~OH p r o d u c t i v i t y and for
caflalyst stabillty. This ~ 9 p t m m a l c o n c e n t r a t i o n is
c r i t i c a l for the o p e r a t i o n ot c o m m e r c i a l r e a c t o r s and
depends on r e a c t o r o p e r a t i n g c o n d i t i o n s .

USE OF BASF S.3.85 CATALYST

General Background
i

In order to c o m p a r a t i v e l y test the c a t a l y t i c activity of BASF S.3.85


catalyst, a series of k i n e t i c experiments were made on the l-liter
slurry reactor. The slurry concentration for this k i n e t i c study was
15 grams catalyst (-140 mesh U.S. Standard Sieve) in 525 cc (STP) of
WJtco-40 oil. It was found in a prior work (Lee, et al., [2]) that
neither external mass transfer nor pore diffusional rate limits the
overall reaction rate at this slurry concentration and with the
crushed catalyst.

Intrinsic Kinetics

Table 3 shows a summary of k i n e t i c data with BASF S.3.85 catalyst


under CO-rich syngas in c o m p a r i s o n with EPJ-25 catalyst. At 236 C
and 940 psia, and n o m i n a l l y the same conditions, the BASF catalyst

3-118
showed higher actlvlty by 3 5 to ii % over the EPJ-25 catalyst
However, it should be noted that the BASF catalyst has a higher
speclfzc surface area than the EPJ-25 catalyst The comparison
reported is not in terms of the zntrlnszc reactzv~ty at the actzve
sztes, but per unzt mass of the catalyst

FABLE 3
KINETIC DATA ON BASF S 3 85 CATALYST

Run Identlflcatlon BOO1 BOO2 BOO3 BOO4 EPJ-25

T, C 236 236 210 250 237


P, Psza 940 940 940 940 940
Exzt Gas ComposltlOn, mole %
i

Hydrogen 21 3 21 9 29 9 21 4 22 8
Carbon Monoxlde 47 4 47 7 47 9 47 i 48 5
Methane I0 8 i0 8 9 6 10 9 I0 1
Carbon Dioxide 9 4 9 4 8 4 9 7 8 9
Water 0 7 0 06 0 08 0 06 0 1
Methanol 10 4 I0 2 4 1 I0 9 9 4
Rate of Reactzon, gmoles/kg cat hr
Hydrogen - 51 3 - 48 7 - 23 5 - 52 5 - 44 5
Carbon Monoxzde - 27 2 - 24 6 - ii 1 - 26 5 - 23 0
Carbon Dloxlde - 1 0 - 0 55 - 0 65 - 0 29 - 0 8
Water 1 59 0 14 0 21 0 13 0 2
Methanol 22 96 22 99 10 41 24 06 22 2
NOTE R u n s BOO1 and BOO2 show the excellent reproduclblllty
of the d a t a

An effort has been made to fzt the reactlvzty data uszng a global
rate expresslon based on a drzvlng force type hydrogen concentration
difference [i], z e ,

rBASF = A exp(-E/RF) (C H2 - C H2,eq) n (5)


where A, E, C H2 and C H2,eq denote the Arrhenzus frequency
factor, actlvatlon energy, H2 concentratlon In the llquzd phase
and the H2 concentratzon at equzllbrlum, respectzvely Table 4
shows all the necessary lnformatzon for the rate expresslon
development As shown zn Figure 3, the least squares method ylelds,
A = 7 4054 x 109 cm 3 ozl / (g cat sec)
E = 25,962 cal/gmol
n : I

and rCH30 H and CA'S are zn gmol/gcat-sec and gmol/cm 3 ,


respectively

3-119
6

c) 5-

C,~ :z:

4
2
L_fO

II.i=

1.9 2.0 2.1

(I/T)x 1000

Figure 3. An A r r h e n i u s T y p e Plot f o r the F i r s t O r d e r


Kinetics with BASF $ 3 . 8 5 C a t a l y s t

3-120
| KINETIC STUDY
TABLE
ON BASF
4
S 3 85 C A T A L Y S T

Run Identzfzcatzon BOO1 BOO2 BOO3 BOO4

T, K 509 509 483 523


P, atm 64 64 64 64
Llquzd Phase Concentratzon zn the R e a c t o r , gmol/l
~ya rdoToong e nM o n o x i d e 0
0
2250
4701
0
0
2280
4686
0
0
2867
4635
0
0
2198
4426
Carbon Dioxide 0 1621 0 1613 0 1518 0 1550
Methane 0 1379 0 1357 0 1195 0 1331
Methanol 0 5393 0 4975 0 2619 0 4462
Water 0 0078 0 0066 0 0139 0 0051
Lzquld Phase Concentratzon at C h e m z c a l Equzlzbrlum, gmol/l
Hydrogen 0 1152 0 1173 0 0529 0 1471
Carbon Monoxlde 0 5564 0 4949 0 4017 0 5465
Carbon Dlo×zde 0 0679 0 0710 0 0785 0 0615
Methane 0 1379 0 1284 0 1133 0 1321
Methanol 0 5471 0 5239 0 4245 0 4357
Water 0 1020 0 1216 0 1121 0 0986
rMeoH gmol/kg cat hr 22 96 22 99 I0 41 24 06
C H2-C H2~eN 0 1098 0 1107 0 2338 0 0727
. . . . . . . . . . . . . . . . . . . . .

rMeOH
209 107 207 678 44 525 330 949

in ......
rMeOH-
.... 5 3428 5 3360 3 7961 5 8020
C

i000 / T, K zT 1 9646 1 9646 2 0704 ---~-9~20---

Morphologlcal Information on B A S F S 3 85 C a t a l y s t

Figures 4 and 5 show the pore szze dzstrzbutzon in BASF S 3 85


catalyst, measured wzth a Mlcromerztlcs Pore Slzer Model 9300 w h i c h
has an intruslon capab111ty of 3 0 , 0 0 0 psz (60 A n g s t r o m s ) As shown
zn ~zgure 4, the most probable pore slze is a p p r o x l m a t e l y 100
Angstroms and the pore volume of the particle is 0 23 cc/g, after
e x c l u d l n g the l n t e r p a r t l c l e v o l d a g e C o m p a r z n g this l n f o r m a t z o n with
EPJ-25 catalyst, the most p r o b a b l e pore szze of BASF c a t a l y s t zs
s m a l l e r (i00 vs 350 A n g s t r o m s ) and the BASF c a t a l y s t does have more
mzcropores than the EPJ c a t a l y s t Also worthy of note is that the
pore volume of
BASF (0 23 cc/g) zs much lower than that of E P J - 2 5
(0 48 cc/g) Thls may be better z n t e r p r e t e d by sayzng that the
poroslty comes more from m a c r o p o r e s whzle the s u r f a c e area comes
l a r g e l y from m z c r o p o r e s

3-121
D I FF. PORE SIZE DISTRIBN.
BRSF 53-85 UNUSED PELLET5

¢.J
LJ

I.t.J o

_j¢=
IO

"4 0
Q

0
e,,

........ I ........ I ' - " ' ..... I ' ...... I ........ I ' ....... I
10 1 z ~ -t
~onE ~ . E T ~ : ~N~RO.~ °' ~°
Figure 4.

3-122
I
l

u3
e=
DIFF. PORE S I ZE D I STR I BN,
BR$F 5 3 - 8 5 UNUSED: -lu=O U5 MESH

¢3

EJ

E~

F
b-~¢3
D

;l

' ' ' ' '"'1 ' ' ''""1


,~,- =,,,~ : ~, ,=,;-'-: -~, =,,,, = ........ , l
Ql tO v

Figure 5.
i i

3-123
Wate~ ~ A n a l y s i s

Informal ion regarding the a m o u n t of w a t e r inside the c a t a l y s t under


process conditions is e x t r e m e l y valuable, since water is also formed
~Js a byproduct and plays an i m p o r t a n t role in e q u i l i b r i u m (both
physical and chemical) as well as in the product diffusion/mass
t:ransfer. In a d d i t i o n , it a p p e a r s that the presence of w a t e r at the
surface of the c a t a l y s t is r e l a t e d to some degree to the degradation
ol~ catalyst structure by metal leaching and/or crystal size growth
b~ hydrothermal synthesis (Lee, et al. [2]; Jerus [II]; Kulik

Therefore, various attempts were made to d e t e r m i n e the a m o u n t of


water contained inside the pores of the c a t a l y s t as well as the
cr.n c e n t r a t i o n of water in the bulk liquid phase under process
cr~nd:itions. The scheme for the d e t e r m i n a t i o n of the s o l u b i l i t y of
wat~r into oil has been published elsewhere (Lee [I]; Ko [13]). As
for the ~ater retained by the solid catalyst, the K a r l - F i s c h e r
tJtratir, n method seems to be most suitable and provides the most
~'produc[l)le data. Using a Karl-Fischer apparatus, the d e t e r m i n a t i o n
of water content in the c a t a l y s t is a two s t a g e job; one being the
~na]ysis of water in the oil and the next being the a n a l y s i s of
water in the oil + c a t a l y s t sample. To be able to m e a s u r e the data
at ~ ~ose to p r o c e s s conditions, the oil and catalyst samples must be
kept i~ sealed bottles under a N2 blanket, and the m e a s u r e m e n t s
n~us[ be n~ade immediately. The data obtained with EPJ-25 and BASF
S.3.85 catalysts are g i v e n in T a b l e 5.

TABLE 5
NOISTURE CONTENT IN USED M E T H A N O L CATALYSTS

EPJ-25 BASF S.3.85


Sample 1 Sample 2 Sample 3 Sample 1 Sample 2 Sample 3

¢ran~s Water 0.0099 0.0100 0.0120 0.0176 0.O771 0.0444


t;ram (iata [yst
Crams ~Ja~er 0.00026 0.00026 0.00025 0.00043 0 00043 0 00043
¢~- ]I]1 OIL] " "

3-124
I
An .<~hu~, ii~ [able 5, the w a t e r content in the BASF catalyst was
:~l~nJfic;~,,Lly higher than that in the EPJ c a t a l y s t . It s h o u l d be
,~<~i_,_~,l [l~dl buLb catalysts were subjected to n o m i n a l l y same process
<~nditions and that the w a t e r production rate (at s t e a d y state) with
I$,,\~F cv~tJLyst :,,as much lower than that with EPJ c a t a l y s t . This
~.~:~os i i,,, f o l l o w i n g possible conjectures:

i) 1,, [e,,,,s of m e t h a n o l s e l e c t i v i t ~ BASF S . 3 . 8 5 may be


a better c a t a l y s t than the EPJ- catalyst.
2) i)ue [,:, the m i c r o p o r e s t r u c t u r e of BASF c a t a l y s t , the
~,~Ler escape mechanism from the pore may De less
e [ fic :i e n t.

3) II~ ~her water concentration in the c a t a l y s t may


:i ,~,] i c dmore hydrg~hillic
Le nature of metallic
,;u~ta]vst than of the el±.

i ul tller a~i~_i]>,sLs of w a t e r content in the used catalyst and the oil


~ ct,rr-~i~tly underway to determine the nature or e x i s t e n c e of
,~iL3]ySt L-,Cl'uolllrTil d e g r a d a t i o n by l e a c h i n g .

N I(:ITOSCOI'I< FIIIL<~ONENA IN METIIANOL S Y N T H E S I S CATALYST

(;(~ 11 t7 i" 71 ] I~,;t ~" k F] r <~ u r i d Ii


Ii~, <omme,cia] ization of the liquid phase methanol synthesis process
~n be t~l<en up with confidence only after the v a r i o u s causes for
i I I
,~t ~]yst Hescl ivmtion have been studied in detail. It is n e c e s s a r y !'i
to est_altl :i s h tl,e e x p e c t e d catalyst life under reactive conditions in
,~i ct~?r t,J detu ,'mine the optimal plant size and profitability.
i
Inv~st [Z,l~ ion of various anticipated causes for catalyst
d('ac t iv,_~ L i i i i-i with regard to the liquid phase process have been in
I~l ~F, I- t' s 5 I-or ,l,ite some time. Thorough research efforts have been
ill<l it @ i_ ii i c~ the c a r b o n y l poisoning problem [14]. Thermal stability of
i
t q
I lit, raLdlyst and the reactor to determine the potential for i

s also been studied (Lee, et al. [15]). Even after


t l,~,sc p,-,,h] euis have been resolved, tile c a t a l y s t still deactivates,
i
I h,>cit4h ~ I uwl)' , due to some unidentified reasons. To p o s i t i v e l y
!
duses for d e a c t i v a t i o n , various parallel studies are
b e i ll'4 ru;~cle , i ,~cluding structural degradation (Lee, et al. [2]),
lIvd r,_,Lbul INd ] S ynLhesis (Jerus Ill]), crystal size growth (Kulik !
i
I lhl) .... i i- b c i l ] ,leposition (Lee, et al. [2]) etc. Some of the o n g o i n g
!
('1 [ ~ - I I L s ;lliti [_lit' ir r e s u l t s are presented here.
!
i I *I

t ,

3-125 !I
X-Ray Analysis of C a t a l y s t Samples

ILach crystal structure has a u n i q u e interatomic spacing, thus each


compound with a defined crystal structure has a unique X-ray
,:l~FfracLion pattern. Even when these compounds are intimately mixed
as t h e > , are in a catalyst, it is p o s s i b l e to s e p a r a t e out the
diffrac[ion effects of each of the c o m p o u n d s .

'['he detern~ination of the c o m p o n e n t s present can be very difficult


v~ithout a d e q u a t e instrumentation and an e a s i l y usable data bank. The
\ , ,, ~',,,,,!,,, Pi~,,:, : ~,,~" ~:~s'".,",':-: .=v.~!?a~:e :z us ~as ~uipped
u. l l l , ,, ,I,,dlcal. ed I'DI'-Ii COlllpU[eI', which executes a search - match
program (run by a s o f t w a r e called "Sandman") that identifies a set
,:,f p o s s i b l e compounds by m a t c h i n g the d i f f r a c t i o n pattern. From this
set the final identification must be done manually. This job is
simplified by a large data bank on the c o m p u t e r and a graphics
software that a l l o w s for a v i s u a l matching of d i f f r a c t i o n patterns.

A large crystal, when X-rayed, gives sharp spikes in the r e s u l t i n g


l:,E~tLern. When a c r y s t a l l i t e of a small size is so X - r a y e d , the sharp
peaks broaden. This broadening can provide information about the
average sizes of crystallites present. One cautionary note, the
analysis assumes that all crystallites have the same shape, which
for practical purposes is not an unreasonable assumption. The
Scherrer formula gives us the size of the c r y s t a l l i t e s , 't'.

t = 0.9 A / B cos eB (6)

where % = X-ra w a v e l e n g t h in A n g s t r o m s
B = w i d t ~ of peak at half N e i g h t in r a d i a n s
0B = Bragg a n g l e in degrees.

T hc radiation used was the Cu K s and this radiation has a doublet.


~':~l r a d i a t i o n has a wavelength of 1.54056 Angstroms and
]. 34435 Angstroms. For peak broadening measurements the
Cu kc~ 2 radiation effects were subtracted out. The instrument
b,-oJderL~F,g was measured using mica sheets and found to be 0.06
degree. This was subtracted out of the m e a s u r e d value of B.

3-126
Phase A1,~ivsis R e s u l t s

Interpretations were made of the following three patterns:

(i~ Pattern EPJOOI.RD. The catalyst samples had been used in the
reactor at 63 atm and 237 C for 80 hours. The feed gas c o m p o s i t i o n
was: H~_ 35.70 %, CO 47.64 %, CH 4 8.48 %, CO 2 8.18 %. The liquid
co~lcentration in the reactor was H 2 0.225 gmol/liter, CO 0.47
gmol/liter and CO 2 0.162 gmol/liter.

From the X-ray patterns, Cu metal was detected in large amounts as


evidenced by very large peaks in F i g u r e 6. The compound CU20 was
also observed, and its small peaks in F i g u r e 6 indicate a relatively
smaller content. Also, a positive indication of the e x i s t e n c e of
ZnCO 3 ~,as found by peak matching, but in small amounts.

12) Pattern EPJOO8.RD. The catalyst sample X-rayed was from the
e×tended run made without CO 2 in the feed gas. The run was made at
237 C and 64 atm for 80 hours and the reactor feed gas c o m p o s i t i o n
was: H2 34.4 %, CO 55.5 %, CH 4 i0.I %. The liquid phase
concentration that the c a t a l y s t had been exposed to was: H2 0.252
g~ol/liter, CO 0.631 gmol/liter, CH 4 0.187 gmol/liter. Again a
strong Cu p r e s e n c e was detected (Figure 7) and CU20 a p p e a r s to be
present in small amounts. However, no indication of ZnCO 3 was
observed. The result shows that patterns EPJOOI.RD and E P J O O 8 . R D are
e~;sentially pretty close to each other, except that no ZnCO 3 was
i,,dicated in EPJOO8.RD.

(~) Pattern EP25VI.RD. This catalyst sample had been exposed to a


CO-f~ee feed gas, (i.e., under a very high partial pressure of
C(~), at 210 to 250 C and 64 atm. The feed composition was H 2
63.3 %, CO 2 31.7 % and CH 4 5 %. The catalyst had been used under
these coI1ditions for 60 hours. As shown in Figure 8, Cu metal was
present in large amounts and Cu20 was present in smaller
quantities. However, most interestingly, ZnCO 3 showed a very
strong presence. This is likely due to the high partial pressure of
CO 2 in the reactor, causing a large portion of ZnO to be c o n v e r t e d
to Z~JCO3. More discussion in this subject is p r e s e n t e d later.

3-127
~IIIW ..................... ",-~'1111JnmllllJ iI -i'll I ......................................... " :~" . . . . . ,li,, - ~ ir~ ~ : - ~'~= :: fq'~rT-'

xlO 3 [ SaT~,~,le: EPJ00 File: EPJOOI,RD 13-MAR-86 10:17

5.00 >,
,m
4.05
r-.
3,20
2,45
I .80 -- ZnCO~I ~ ZnC03
1.25
0.80
0.45
0,20
0.05

CO
I
10,0 15.0 20,0 25.0 30.0 35.0 40.0 45.0
PO
5.00
CO
2 x Bragg angle
4.05
3.20
2.45
I .80 Cu Cu
I .25
0.80
0.45
0.20
0.05
I I ! ! J ; I ! ! ! I I J 4

45.0 50.0 55.0 60.0 65.0 70.0 75.0 80.0

F i g u r e 6. X - R a y Diffraction Patterns

i ii ,, ,,,, , - , - . . . . . . . . . . . . . . . .
xlO 4 I Sam~,le: EF'J00 File: EPJOOS.R[~ 13-MAR-86 10:53

I ,00
0.81 u)

0,64 ¢=
C u

0,49 ¢-
overlap tl
m
0.36
0.25
0.16
0,09 ............ " "''ca-,=---'# ~-- "

0.04 ,~,~,=,~.~..-: ._ ..- . . . . . . . . . . . . . . . . . . . . . . . .


0,01
Co
i 10.0 15.0 20.0 25.0 30.0 35.0 40.0 45.0
PO
,,.(3
;t .00 2 x Bragg angle
0.81
0.64
0.49
0.36
Cu Cu
0.25
~ , / ~ ~ . A1203 .-.. ~ . ~ = ~
0.16 • .,--~..::,---~,_~.:..,..,. " _ _.--- .-

0,09
0.04
0.01
a i i | i J J i ! i ! J i

45.0 50.0 55.0 60.0 65.0 70.0 75.0 80.0


Figure 7. X - R a y D i f f r a c t i o n P a t t e r n s
I Sample: e p 2 5 v File: EP25UI.RD 20-FEB-86 14:36
xi0 3 [
5.00 >, Cu
4.50
4.00 c 4-
3.50
C
3.00 --
¢o ¢~ 0 0 < O
2.50 c
O O N ,,, 0
2.00 o _~ ~. ,- j i

1.50
1.00
k.
0.50 .--..~. ~- ". '--'- "=-~ "=--"":...... "-" :"-- "=---'~ I I I I I I I
= • • i i • i
l

!
10.0 15.0 20.0 25.0 30.0 35.0 40.0 45.0
5.00 2 x Bragg angle

4.50
4.00
3.50 ¢0
3.00 ¢0 O ¢o ~ ¢o
O O O O
2.50 00 Cu c od 0 o~. Cu
2.00 ¢: Jk N ~ ~ N A
1.50
1,00
0.50 II , I I , II , I , I I , I . . .I .I . I I I I , I , I , ,

45.0 50.0 55.0 60.0 65.0 70.0 75.0 80.0

Figure 8. X-Ray Diffraction Patterns


I
I S~rnple: epj25 File: EPJ250.RD 2 ~J-FE~-~_:~ 14 :;~:~
~"10 ~ I
5.0~3
.50
03
.00
.50 ¢-.

.00 -- CuO CuO


2.50
2.00
1.50
1.00
.50
~)1~--- i
' ~ - ~ " - ~= : : - -
6
I =
I I II I I I
!
10.0 - 15.0 20,0 25.0 30.0 35.0 40.0 45.0
tu
5.00 2 x Bragg angle
4.50
# .00
3.50
3.00
£ .50
,--'.0o / CuO CuO

1.00

0.50 | I I I I I I I I I
- - I I i I I I ! I I I I I I I

45.0 50.0 55.0 60.0 65.0 70.0 75.0 80.0

Figure 9. X-Ray Diffraction Patterns (Fresh Catalyst)


(4) Pattern EPJ250.RD. The unused, unreduced, fresh EPJ-25 catalyst
was X-rayed in o r d e r to c o m p a r e the r e s u l t s . The pattern is s h o w n in
P:i s u r e 9.

X-ra i Crystallography on C o p p e r Compounds

It was confirmed that in an a c t i v e catalyst of the E P J - s e r i e s , both


Cu + and Cu ° exist as c l e a r l y seen in F i g u r e s 6, 7 and 8.

It ~as also found that the reduced, active EPJ c a t a l y s t does not
contain any CuO, i.e., Cu ++. However, it was not possible to
,]eLect any evidence of catalyst overreduction when used in a
C02-free environment.

X-,4y Cr>s1:allography on Zinc Compounds

llril~ublJshed reports have said that the role of CO 2 in m e t h a n o l


svnthesis is to stabilize the c a t a l y s t by f o r m i n g Z n C O 3 from ZnO.
Analyzing the o b s e r v a t i o n s more scientifically, we have reached the
l!ol]o~ing conclusions:
(i) ZnCO 3 was found in some of the used c a t a l y s t s .
However, this catalyst was s u b j e c t e d to e x t r e m e z y
high partial pressure of CO 2 without CO in the
syngas mlxture.
(2) All other catalyst s a m p l e s w h i c h did not c o n t a i n a
large a m o u n t of Z n C O ~ had been s u b j e c t e d to normal
svngas conditions where CO 2 p a r t z a l pressure was
not h i g h .
(3) Therefore~ the formation of ZnCO~ by a reaction
between ZnO and COo was c o n f i r m e d uffder hiRh COo
partial pressure renditions. The ZnCO~ ~ormation
under low COo partial pressures does ~ not seem
~ o s s i b l e b a s e d = o n the t h e r m o d y n a m i c c a l c u l a t i o n given
Je 1 ow,

To c o n f i r m the c l a i m (3) thermodynamically, the e q u i l i b r i u m constant


l!or the reaction, ZnO + CO 2 = ZnC03, was calculated.
Tl~ermodynamic data published for ZnO, C02, and Z n C O 3 are shown
Jn ]'able 6.

IJsing [he published data at 25 C, and i n c l u d i n g the t e m p e r a t u r e


dependence at 237 C, the equilibrium constant for the reaction,
Zr,O + CO 2 = Z n C O 3 b e c o m e s :
at 25 C, Ka = 2015
at 237 C, Ka = 0.014

3-132
TABLE 6
THERMODYNAMIC DATA

p Gfo .fo s°
Spec_i es
C atm kJ/gmol kJ/gmol Jlgmol K J/gmol K

ZnO(cr) 25 1 -318•30 -348.28 43•64 40.25


ZnCO~(cr) >3 1 -731 52 -812 78 82 40 79.71
CO,~ ( £ ) "5 I -394 359 -393 509 213 74 37 I I

Cp,¢(,2 = 2 6 . 0 1 7 + (43.526 x 10 -3 ) T - ( 1 4 . 8 4 2 x 10 -6 ) T 2

$ourci: The NBS T a b l e of C h e m i c a l T h e r m o d y n a m i c P r o p e r t i e s .


J.Phys. & Chem• Ref. Data, Vol.ll, S u p p l e m e n t 2, 1982

A:~ shown from the value of the e q u i l i b r i u m constant, the ZnCO 3


l!,,rm;:~tion r e a c t i o n proceeds in f o r w a r d direction only w h e n the CO 2
part ia] pressure is very high.

I)ue to the extrapolatory nature of the v a l u e of K a at 237 C, any


,,,,meri,:al [,rediction of the partial pressure of CO 2 r e q u i r e d to
nl,ke the forward reaction proceed is a v o i d e d here• Therefore, it is
a s~[e beL that the formation of Z n C O 3 is not likely to be a
significant reaction with high-H 2 or high-CO syngas feed in the
I iq,,id Phase Methanol Process, unless the CO 2 p a r t i a l pressure is
v~r!/ l,ig],. However, the ZnCO 3 formation reaction is b e l i e v e d to
i:;~k, l~lacc: when CO2-rich syngas is used, as is the case w i t h some
Lur,,l>,~ ~ i~ocesses. It s h o u l d be n o t e d that the c a l c u l a t i o n results
;,,e ,:c,,isisLent with the observed X-ray crystallography data, as
, v~,Ivnccd by F i g u r e s 8 and 9.

( ~ v v ~ , v l [ite $iize M e a s u r e m e n t

iliL~i:e ; l a v e been reports and articles (Natta [17]; Kulik [16])


,-cg~,rdlrlg the strong correlation between the r e a c t i v i t y and the
'-,'I'I" '- c r s s t a l l i t e size. As the c r y s t a l l i t e size grows, the a c t i v i t y
,>i- Lhe catalyst is reported to decrease• In catalysis, this
l~;Je4,,,,uc~on is often classified as " a g e i n g " of the m e t a l l i c catalyst
(Pausescu, eL a l • [18])•

3-133
Comparing the crystallite sizes of EPJ-25 catalysts between the
"properly used" and the "misused", an interesting observation was
~J,,,I., . I,. I 1.1,, I"ll'~'f , I ll~, I , , I i. "l}r,)l),'l I y II~,,d" refers to t h e catalyst
used under normal syngas conditions for 80 hours, whereas the term
"misused" refers to the catalyst used under rather abnormal (though
intentional) conditions, i.e., without CO 2 (as mentioned before)
for 80 hours. As explained earlier in this paper, the reactivity of
the misused catalyst was 80 % of the original reactivity. However,
the two catalysts show practically the same crystallite sizes, viz,
195 .$2 Angstroms for "properly used", 198.76 Angstroms for
"~r~isused", even though the activities of the two catalysts were
quite dJ#ferent (Refer to Table 7).

TABLE 7

CRYSTALLITE SIZE MEASUREMENT

0
Sample %, A @B deg B Average Crystallite
m
radian Size (angstroms)

EPJ@OI.RD 1.54056 21.65 0.43 198.76


EPJOO8.RD 1.54056 21.69 0.43 198.82

This experimental fact implies that the reduced activity of the


catalyst may have been due to totally different reasons, such as
carbon deposition mentioned earlier. Also implied by the result is
that, since practically the same number of hours have been applied
to the two catalyst samples under process conditions, the degree or
level of "ageing" the catalyst may have been identical. Therefore,
it can be suggested that the crystallite size is more of an
indicator of "age", than of deactivation.

Analysis of Used 0ii

The possibility of metallic ingredients leaching into the process


oil during operation was first reported with preliminary data on
atomic absorption spectroscopy (Lee, et al. [2]). To refine the
measurement procedure, various methods have been tested in terms of
reproducibility and accuracy. The methods tested include: (a) direct
measurement, (b) ashing, (c) wet ashing using strong acids and (d)

3-134
asl,i,lr, wJtl, oxidizing agent. Among the methods investigated, the
cunce~,trating technique based on ashing with an oxidizing agent
MgNO 3 ~,as the most reliable one, showing an average recovery level
,~f 80 Z ,,[ ~ h e original metal content.

The ,,~ I s,,mples analyzed were collected from the slurry reactor
ailLer ](I0 l~ours of operation with EPJ-25 catalysts and filtered
th,,ro,J~hLy to make sure that no suspended particle is present. The
ai:,-,mi~ ,~bsorption spectroscopy data show that the oil contains 6-20
ppm ,,I At. Considering the low slurry concentration (20 g catalyst
p,-r ~,i,() cc ('3TF') of oil), the concentration of A1 may be high enough
tu I,,: ,,[ cot,tern. Even if all the present A1 dissolve into the oil,
]In ,,,~ce,~l ,r, tion to be detected by atomic absorption spectroscopy
~,,,,~],i i,,: o,,Iv 0.2 % or 2000 ppm. Therefore, the leached level of A1
:i. tl,e o:i i accounts for approximately I % of the total A1 in the
c~tal,,:~(, Lf 21[) ppm is detected.

Si~,cc tile l u v e l of metallic leaching into oil differs from sample to


s,JrupI, , depunding on the mode of operation, the type of catalyst,
th,, n,lmber of hours, the type of syngas, etc., any drastic
(,~ncl,~:~J,,n Js avoided here. llowever, if the problem is found to be
seri,,us _,~L~:~- a s u f f i c i e n t l y long commercial or simulated run, some
means ,:~f p~eventing this would be necessary. Based on the AA
an.,]y~i,: ,:,~, , r L i f i c i a l samples prepared from an aqueous solution, it
,~i,l~en~ tl~,L the presence of water inside the catalyst accelerates
su,:ll ,I~ ! , , i ~ L , J a L i o n of catalyst support [2].

i',,~,_, ,,~, ,I, t r i b u t i o n s o[" the used and unused EPJ-25 catalysts were
mu L~U,L,i ~,,, iiicromeritJcs Pore Sizer 9300, as shown in Figures i0
a,,,l I I, II was fouud that the used catalyst has a higher average
p,,ru s,.'u i i,a,~ the u.used catalyst and the specific surface area of
the uf,u:~::,l c ~talyst is higher than that of the used catalyst.

II iJl~I, ic,s that the catalyst morphological structure also changes


dHr I II}' I h~2 process of operation. In addition, the reduced surface
cl I- t' ,3 V I LI C' J s to be at least partly blamed for the reduced overall
ac/ivil

3-135
'z. DIFF. PORE SIZE DISTRIBN.
Unused Cetil~yst EPJ-2$ P e J i e t 8

1
o

E
%.
~o
~o
,e
o
@
E
m0

Oq,
>o
,e
o
oo

emo
mq

o
,e
o

o
o
i i i ilmll|
o ' ''''"I ' '''''I ' '''""I "-'-'''"I- r-'''""l
108 108 '04 '0' '0' 10 7
Po re D t a m e t e r : Angstroms

Figure 10.

3-136
O
m
DIFF. PORE SIZE DISTRIBN.
Deacl Jvated C a l a l y s t EPJ-25 P e l l e t s

~m
(3o
(D
E
N,,I
O~.
:>o
¢;-

(3N
O
¢;- P

O
0

I0 1~20 v0 = 10 ~ IO s !01 10 7
re Diameter: Angst r0ms

Figure 1 1.

3-137
C()17c-'i. lJS] ONS

l'l~c~ fo] ]ot, i n g conclusions can be drawn from the current

(I) ~Jelll-,c~riol p r o d u c t i v i t y drops rapidly in the absence of c a r b o n


diuxJde wl,ei-i s y n g a s with a high carbon monoxide to h y d r o g e n ratio is
used ,

(2) ~.rit i, high CO partial pressures and without CO 2, carbon


,iu p o s i l _ i , ~ r l is possible and it was evidenced by d e t e c t i n g CO 2 in
tlJe ver, i_ 9,as m i x t u r e .

(i:) ITl-,c,J, 112 + CO 2 is used as syngas feed without CO, methanol


I>ru,:l,sctJ,-,n proceeds mainly via a two stage scheme, i.e., reverse
water ~s shift reaction followed by c a r b o n monoxide hydrogenation.
']'h:is obsorvation was at least true in macroscopic sense, when
,_'ml-, ! o y i n , , t vl-,e m a t e r i a l balance concept.

(4) 17hen H2 + CO is used without C02, the methanol production


r~:,i:e l:,econ~,~s s i g n i f i c a n t l y reduced. Immediately after reverting to
,,orr~.~l svr, gas with 8 % C02, the m e t h a n o l rate increases rapidly,
[Ik:~ugh n,Jt to the original level. No direct experimental evidence
w~,s found to s u p p o r t the possibility of c a t a l y s t overreduction as a
c;Juse fo, the reduced activity in a C 0 2 - f r e e environment.

(;) Llr,d+i'r high partial pressures of C02, ZnO converts to ZnC03.


]I,i.~ ~,~ positively identified by X - r a y crystallography. The same
pl,<,,,,mer,,:.,i was not found in o t h e r catalyst samples which were used
,_,J~!~ u,-,,t,,c low CO 2 partial pressures. The experimental results
~.,,.,-e in di.+reement w i t h theoretical thermodynamic calculations.

(l~) It w,~,s reconfirmed that AI203 support in EPJ catalyst is


:~,,I,iect ~,, leachin~y into the process oil (Witco 40 or F r e e z e n e 100).
]I,:,t _ipp, ,rs to be m o r e conspicuous in the presence of w a t e r .

3-138
(7
B,ISIT S.3.85 catalyst is less porous than EPJ-25 catalyst;
howev{?r BASF
catalyst has more micropore structure than EPJ-25
C ,:~ La 1 v :--, I_ thus
providing more internal specific surface area. The
reacLi~ [ty (at
237 C, CO-rich syngas) per kilogram of catalyst was
3.9 t.:, Ii % higher with the BASF catalyst. A rate expression for
UASF ,:aLalyst in liquid phase methanol synthesis is proposed in this
l,~Iper. The developed kinetics show that the BASF catalyst behavior
is m,re sensitive to the temperature change than the EPJ catalysts,
evldenced by its higher activation energy.

H~,re work is being done in the area of (a) role of CO 2 and H20
iJd meLl~anol synthesis, (b) mass transfer, and (c) causes for
c ~lalyst deactivation. Other efforts are being made to improve the
!,rocess Jn terms of chemistry, engineering design and understanding,
,~nd pr,,,_ess opLimization.

A C F NOW I, 1:: DI3 I'IHENTS

II,,, auth,:,rs would like to express their gratitude to Dr Volker M


Sc lJw e ll d @lli~ N " •
of BASF Wyandotte Corporation, and Dr. Paul Jerus of
IInJ~ed Catalyst, Inc., for having provided us with the samples of
used in this work. The authors also thank Dr. James K.
I1,:~ I d y o E
the Department of Chemistry and Dr. Annabelle Foos of the
I), p,~rLmL, jlL
of Geology, The University of Akron, for their guidance
n f, d he t p
in carrying out the analysis of the various samples. The
,3k~G I S L,~II(Ze
rendered by Mr. M. E. Polasky, Mr. K. Rodrigues, Mr. p.
k'~ r t,h,-J r i ,
and blr. D. Hayes in completing this work is also
g l J [ (2l-U I J.y ,~cknowledged.

3-139

4
R EF 1717I;N(GILS
]. ::i. R e s e a r c h to Support Deyelopment of the Liquid-Phase
~1 I-~ L I-i El ~,ol ~vntbes]s Vrocess, E F R I AV-4479, Electric vower
,<I, lhstltute, Valo Alto, CA, February 1986.
I e {7: , ~i., If. Ko, A. Sawant V. Parameswaran, and T. Sullivan,
" I< I±~ _ ~,-ch into
_ $ Im' Nass Transfer, Oil and Catalyst Degradation in
' q I I -i ~i Fhase Nethanol', Proceedings: Tenth Annual EPRI
l;,ijiTi t r a z r o r s ' Conference on Ckean Llquld and SOild Fuels, EPRI
,t H - 4 Z , b - S i t : , O c t o b e r i~J85.
3. Ll~ec , I<., Adv. Catal., 31, 243 (1982)
4. 7, a ::tan , Yu. B., A. Ya. Rozovskii, L. G. Liberov, E. V.
'" ] :i " i l i S ] : i G. i. Lin, S. M. Loktev, and A. N. Bashkirov,
,\ k a ,i i! :~i u k ', SSSR 224, ]081 (1975).
') .
Yu. B., A. Ya. Rozovskii, G. I. Lin, E. V. Slivinskii,
ev , h. G. Liberov and A. N. Bashkirov,
1, ili~-2i , l<al: al., 16, 809 (1975).
! ,0 !~ ,fl II , Y i] , B,, G. I. Lin, A. Ya. Rozovskii, S. M. Loktev, E. V.
:';] i v i , , a k i i , A. N. Bashkirov, I. P. Naumov, I. K Khludenev, S.
.~ , '/ll~l i i-IOV and Yu. I. Golovkin, Kinet. Katal , i7, 440, (1976).
7. ',~c,;-ov ~l-;Jj , A. Ya. Yu. B. Kagan~ G. I. Lin, E $. Slivinskii,
!-;. N, L o k t ev, L. ~. Liberov.)an~ A N. Bashkiro
killel: , }<at al , 16, 810 (1975 .
3, ~, {I 7 I] I-' t $1hV , V. D F. S. Shub and M. I. Temkin,
]<,net . );;-!t a]., i~, 932 ( 1 9 8 2 ) .
[, ili]!-,, , I I . II ., G. Liu and D Willcox. Amer. Chem. Soc.,
F'ueI I-li v J s ion Preprints, i9 (5), 194 (1984).
O. t~rc~ n lJ I' C , T . and A. Savari, "Chemistry and Uses of Carbon
l iiu:.; i,le' , PRI RP-2563-5, Final Report, October 1985.
]. leru~ , F., P rivate Communications, 1985.
]2. l,ulJT, ¢. J., Private Communications, 1985.
i '-~,, /o, If., "Thermodynamics of The Liquid Phase Methanol Synthesis
Pr~,c:ss", N. S. hesis, The University of Akron, 1985.
,4 , "-,
:,~->wl, D. ~<i_ T. H. Hsieung, P. Rao, and M. I. Greene
(at,~v<t ~ivity and Life in Liquid Phase Methanol ,
l'rocee~ILrigs: Tenth Annual EPRI Contractors' Conference on
I;I~>aT~ Liquid arid Solzd Fuels, EPRi AP 4253-5k, OctoDer i985.
I :~. I~c,e, 6., J. H. Betty, H. L. Greene,,, S. Desirazu, M. Ko, V.
!'~r ~me-:~:Tran, and A. Sawant, Thermodynamics, Kinetics, and
ll,ecmal Stability of Liquid Phase Methanol Synthesis",
l'<.,c, e~lings: Ninth Annual EPRI Contractors' Conzerence on Coal
L~quekaction, EFI<I AF-J~Z~-51(, Narcll 198~.
i[.. KulJI<, C. J., EPRI Journal, pp. 48-49, November 1985.
17. ~8 hal.
~ ' ~ aLet
t tG.
a, Synthesis
349 (1955).of Nethanol, Catalysis, Vol. 3,

Ii~, Pau>,~scu, P., R. Manaila, and N. Popescu, and E. Jijovici,


I. AI,I,I. Cryst., 7, 281 (1974).

3-140
PROJECT DEVELOPMEI~T OF A COMMERCIAL COAL-OIL COPROCESSING PLANT

R. H. Shannon

Ontario-Ohio Synthetic Fuels, Inc.


PROJECT DEVELOPMENT OF A COMMERCIAL COAL-OIL COPROCESSING PLANT

R. H. Shannon

Ontario-Ohio Synthetic Fuels, Inc.

Yesterday the representative of Exxon Corporation, in a discussion on the effects of


moisture, ash, etc., on the economics of coal liquefaction said i f i t was possible
to;
"get a coal w i t h low m o i s t u r e and ash, and a high
H/C r a t i o and have an i n f r a s t r u c t u r e a v a i l a b l e ,
you might develop a coal liquefaction p r o j e c t . "

Well, we have that and more. We have;

- low ash, low m o i s t u r e , low oxygen, low priced coal w i t h high H/C,
vitr~inite and v o l a t i l i t y ratings and an available i n f r a s t r u c t u r e .

- Coal conversions c o n s i s t e n t l y in the mid 90's compared to the


conversions in that paper of 75%.

l i q u i d y i e l d s in the mid 70's compared to the 50% noted in t h a t


paper, that is 5 barrels of liquids per ton of coal up from 3 to 4.

- cut capital costs in half per unit of capacity.

- substantially reduced o p e r a t i n g cost p a r t i c u l a r l y hydrogen and


catalysts.

LET ME TELL YOU ABOUT THE PROJECT.

The project is a PROTOTYPE COMMERCIAL COAL-OIL COPROCESSING PLANT to


produce 12,250 barrels per day of light clean fuels. These are good
naphthas, d i s t i l l a t e s and LPG's for refiners. For u t i l i t i e s they can
be blended d i f f e r e n t l y and can be used as very low s u l f u r fuel o i l s
or turbine fuels. Virtually all sulfur and nitrogen was extracted.
In commercial operations these would be recovered as valuable
industrial products.

- The cost is estimated at $225 m i l l i o n .

I t is to be sited in Ohio in an e x i s t i n g i n f r a s t r u c t u r e and at a


confluence of natural gas, crude oil and product pipelines and r a i l
and road services from coal fields 100 miles away.

It is based on HRI's commercial H-Oil technology and the H-Coal and


Wilsonville Advanced Coal Liquefaction demonstration plants.

3-143
I t would use about equal amounts of high s u l f u r and high n i t r o g e n
coals and heavy o i l s for a total of 1500 tons per day.

I t is planned for commitment in 1987, a f t e r a P i l o t Demonstration Run


at large scale and f o r s t a r t up in 1990 to obtain bulk samples f o r
customer negotiations and for precise design data.

OUR STATUS IS THAT WE HAVE;

Completed a major stage of development w i t h over 400 micro and


autoclave laboratory tests in four laboratories.

Completed four extended p i l o t p l a n t runs in two p i l o t p l a n t s , and


many supporting studies.

The p i l o t i n g was in the ebullated p i l o t facilities of HRI, Inc., at Lawrenceville,


N. J. and the Kerr-McGee Corporation f a c i l i t y at Oklahoma. Both are two stages,
closed coupled. The Kerr-McGee unit was integrated with a C r i t i c a l Solvent Deashing
step.

OTHER COMPOHENTS IN THE PROJECT DEVELOPMENTARE;


The State of Ohio has provided assistance over the past two years.
The Ohio Department of Development has responded to our c u r r e n t
applicationto be a financial cosponsor of the project and has made a

"commitment of support of up to $I0 m i l l i o n "

I believe we can f u l f i l l the two conditions required.

EPRI assisted in the development of the technology d i r e c t l y and


i n d i r e c t l y . When Ontario Ohio Synfuels received i t s f i r s t grant from
Ohio, in 1984, I approached EPRI for additional funds. EPRI couldn't
help f i n a n c i a l l y but phoned Kerr-McGee who had some time available in
the p i l o t plant. By the weekend we had 25 drums of Cold Lake bitumen
and I0 tons of Ohio coal in dedicated trucks enroute to Oklahoma.

In e i g h t years on development on coprocessing, out w i t h a t i n cup


being piddled on, these were my f i r s t breaks and I dearly appreciated
them.

Besides Ohio we have a l e t t e r of commitment from EPRI f o r f u r t h e r


assistance.

The City of Warren Ohio has offered the serviced s i t e I described and
has offered a loan on favorable terms of $2.5 m i l l i o n .

We have negotiated a proforma lease with an international equipment


vendor for up to $I00 m i l l i o n at a rate of 6.9%.

The 200 acre s i t e has been optioned. Besides being at a confluence


of o i l , products, coal and n a t u r a l gas d e l i v e r y systems, i t is
serviced with power, r a i l sidings, sewers, water and a 200,000 sq f t
building.

3-144
I 65 m i l l i o n tons of coal have been dedicated to the p r o j e c t . This
coal has been proven e x c e p t i o n a l l y amenable to conversion. On the
basis of conclusive demonstration in the 400 lab tests and in 4 p i l o t
p l a n t runs, t h i s is e q u i v a l e n t to 300,000,000 b a r r e l s of o i l . This
is more than is l e f t in the North Sea.

The coal was subjected to an extensive p e t r o g r a p h i c a n a l y s i s and


compared with coals of known liquefaction behavior such as l l l i n o i s ,
Kentucky and Wyoming coals, and we know why i t is e x c e p t i o n a l l y
amenable.

We have an o u t s t a n d i n g team of h i g h l y q u a l i f i e d personnel whose


experience and c a p a b i l i t i e s encompass a l l the a d m i n i s t r a t i v e ,
process, e n g i n e e r i n g , c o n s t r u c t i o n and maintenance s k i l l s required.
This is the team of Ontario Ohio Synfuels and i t s a f f i l i a t e Ohio
Ontario Clean Fuels, HRI, Inc. and Stearns Catalytic.

We have begun market negotiations.+ The market in our area is 30 fold


what we w i l l produce in the f i r s t plant, i t s a case of maximizing net
backs.

IN RESPECT TO ECONOMICS

We look at many facets of the economics, l e t me describe three;

i) Comparative economics with other sources of o i l

2) Internal rate of return as a project

3) Comparisons w i t h other systems to remove NOx and SOx f o r acid r a i n


control

Let me deal with #I f i r s t , costs of new o i l supplies. New supplies are required to
replace the d i m i n i s h i n g reserves of l i g h t oil in the Western hemisphere and in
Europe. There has been no major find of l i g h t o i l since the Beaufort find 25 years
ago, in spite of expenditures of many b i l l i o n s .

For the l a s t 15 years consumption of l i g h t oil in North America has been n e a r l y


double t h a t of the f i n d s of l i g h t oil. The major s u p p l i e r s to the USA are also
facing declines, these are Canada, Venezuela and the North Sea. Mexico's production
of l i g h t o i l might not decline but i t s internal growth is high, which may l i m i t its
exports of that quality.

I t is recognized that new sources of l i g h t o i l must be developed w i t h i n the western


hemisphere. Offshore d r i l l i n g is taking place o f f a l l shores. Efforts are being
made to bring in l i g h t oil from enhanced heavy oil recovery and upgrading, tar sand
and other NON Conventional sources.

3-145
A COMPARATIVE ANALYSlS WITH THESE OTHER SOURCES SHOWSTHAT;

Coprocessing is more than competitive with resid upgrading which is


commercially practiced.

The p r o j e c t r e q u i r e s less c a p i t a l i n v e s t m e n t , about 50% l e s s , per


barrel of capacity than the f r o n t i e r , enhanced heavy o i l recovery and
upgrading, t a r sands, coal l i q u e f a c t i o n ( w i t h r e c y c l e ) and shale
projects.

The "finding" costs are minuscule compared with the $6 per barrel or
more for "finding" costs for conventional o i l . I t also compares very
favorably with the cost of acquisition by " d r i l l i n g on Wall Street".
Chevron and other o i l companies have acquired o i l at costs in excess
of $6 per b a r r e l . This is more than our " f i n d i n g " and " o p e r a t i n g "
costs.

"Royalty" costs on coal in terms of o i l e q u i v a l e n t are less than 5%


of r o y a l t y c o s t s on c o n v e n t i o n a l o i l as imposed by some
jurisdictions.

- Operating costs for conversion are no more than the operating costs
of an off-shore platform or an enhanced heavy o i l recovery operation,
let alone the additional costs of upgrading heavy o i l .

We are not at a l l e c o n o m i c a l l y disadvantaged to any of the other sources of new


l i g h t o i l e s s e n t i a l to North American industry, industrial users and the
transportation and u t i l i t y industry.

IN RESPECT TO THE INTERNAL RATE OF RETURN;

A very key p o i n t is t h a t our feedstock costs d e c l i n e w i t h the general d e c l i n e in


energy prices. This allows us to maintain a positive margin at low o i l prices. Our
feedstocks have declined more than the p r i c e of our products or the general price
decline of premium low sulfur low nitrogen l i g h t o i l s .

A p r o j e c t has to be f l e x i b l e to c o n t r o l margins, and feedstocks and t h a t s e x a c t l y


what coprocessing is able to do.

COST EFFECTIVENESS IN REMOVING NOX AND SOX

Due to the time allowed me I w i l l b r i e f l y describe only one other facet, that is the
cost of removal of NOx and SOx.

We buy s u l f u r and n i t r o g e n in coal at $35 per ton. We s e l l them at w e l l over $100


per ton. I t does not cost t h a t much to capture them. We get a s i g n i f i c a n t credit
per barrel by the sale of elemental s u l f u r and ammonia.

3-146
In respect to the cost of scrubbers, the M i n i s t e r of Energy of the UK, and many
others simply say they are not the answer in s p i t e of t h e i r very high cost of
installation and o p e r a t i o n . I am t o l d they do not remove n i t r o g e n , and burning
controls only remove up to 25%.

What is the cost of scrubbing sulfur and nitrogen from coal and residual o i l s which
have over 3% s u l f u r and about 2% n i t r o g e n to the l e v e l s of 0.3 wt % r e s p e c t i v e l y ,
which we have conclusively demonstrated? Many j u r i s d i c t i o n s have legislated against
the n i t r o g e n emissions from a u t o m o b i l e s , presumably t h i s will be extended to
nitrogen emissions from other sources such as those from power plants.

Another aspect of coprocessing is t h a t i t produces a s t o r a b l e l i q u i d fuel. The


coprocessor producing t u r b i n e f u e l s f o r a combined c y c l e , t h a t is a COPROCESSOR
COMBINED CYCLE, (CC) would seem to have advantages worth considering along with the
INTEGRATED GASIFIER COMBINED CYCLE, (IGCC). The coprocessor does not have to be
integrated, i t is f l e x i b l e , I don't b e l i e v e the l i q u i d fuel is anymore expensive
than synthesis gas. It certainly removes NOx and SOx in l i k e manner t o t h e
Integrated Gasifier Combined Cycle.

I feel there may be a market there, in any event there are other markets.

I did not come to coprocessing from the coal side. In western Canadian enhanced
heavy operations, f i r e floods, steam floods, oxygen enrichment, miscible floods were
being t r i e d , all were expensive, a lot were f a i l u r e s , most were o p e r a t i n g
nightmares,

I obse~-ved that the heavy o i l t h a t was produced was being p i p e l i n e d over coal
fields. I said Co myself, "I bet the reactor won't know the difference i f there is
a b i t of coal in the o i l , and the cost of the o i l would be averaged down." The BTU
carbon in the coal was less than 20% of the carbon in heavy o i l , and is s t i l l much
less. We tried a couple of pails in 1976 and i t made sense and i t s t i l l does.

The same economic principles apply today. Coprocessing is economically a t t r a c t i v e


and technicdlly feasible.

3-147